Megan Dissertation - Regression Models
Megan
2025-02-21
Manually set working directory to source file location (Session Tab)
library(readxl)## Warning: package 'readxl' was built under R version 4.1.3library(dplyr)## Warning: package 'dplyr' was built under R version 4.1.3##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(tidyverse)## Warning: package 'tidyverse' was built under R version 4.1.3## Warning: package 'tibble' was built under R version 4.1.3## Warning: package 'tidyr' was built under R version 4.1.3## Warning: package 'readr' was built under R version 4.1.3## Warning: package 'purrr' was built under R version 4.1.3## Warning: package 'stringr' was built under R version 4.1.3## Warning: package 'forcats' was built under R version 4.1.3## Warning: package 'lubridate' was built under R version 4.1.3## -- Attaching core tidyverse packages ------------------------ tidyverse 2.0.0 --
## v forcats 1.0.0 v readr 2.1.4
## v ggplot2 3.4.4 v stringr 1.5.0
## v lubridate 1.9.2 v tibble 3.2.1
## v purrr 1.0.1 v tidyr 1.3.0## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
## i Use the ]8;;http://conflicted.r-lib.org/conflicted package]8;; to force all conflicts to become errorslibrary(ggplot2)
library(broom)## Warning: package 'broom' was built under R version 4.1.3TO DO -
REDO INTERPRETATION SECTION WITH UDPATED RACE AND MS SUBTYPE COLUMNS
- maybe excel: sig or no?
Home videos - Log T25FW still not super linear… -> drop outliers?
Home videos - re-do with improved vel pixel proxy metrics
Load and format In-Person Video Data
All videos included in analysis. Videos were included based on if a walking segment could be identified. Some participants may have both a fast walk and preferred walking speed video. Some may have just fast walk or just preferred walking speed.
Each row = 1 video
Preferred walking speed
# Preferred Walking Speed
zeno_pws_df <- read.csv("C:/Users/mmccu/Box/MM_Personal/5_Projects/BoveLab/3_Data_and_Code/gait_bw_zeno_home_analysis/004/000_merged_cleaned_data/zv_bw_merged_gait_vertical_PWS_1_clean.csv")
table(zeno_pws_df$task_pose)##
## gait_vertical_PWS_1
## 217# assign levels to categorical variables
table(zeno_pws_df$race_ethnicity_clean)##
## Asian Black Or African American Hispanic or Latino
## 17 14 20
## Other/Unknown/Declined White Not Hispanic
## 18 148zeno_pws_df$race_ethnicity_clean <- factor(zeno_pws_df$race_ethnicity_clean,
levels = c('Asian',
'Black Or African American',
'Hispanic or Latino',
'White Not Hispanic',
'Other/Unknown/Declined'),
ordered = FALSE)
print(levels(zeno_pws_df$race_ethnicity_clean))## [1] "Asian" "Black Or African American"
## [3] "Hispanic or Latino" "White Not Hispanic"
## [5] "Other/Unknown/Declined"table(zeno_pws_df$ms_dx_condensed)##
## MS, Subtype Not Specified Progressive MS RRMS
## 2 37 178zeno_pws_df$ms_dx_condensed <- factor(zeno_pws_df$ms_dx_condensed,
levels = c('RRMS',
'Progressive MS',
'MS, Subtype Not Specified'),
ordered = FALSE)
print(levels(zeno_pws_df$ms_dx_condensed))## [1] "RRMS" "Progressive MS"
## [3] "MS, Subtype Not Specified"table(zeno_pws_df$clean_sex)##
## Female Male Non-Binary
## 162 53 2zeno_pws_df$clean_sex <- factor(zeno_pws_df$clean_sex,
levels = c('Female',
'Male',
'Non-Binary'),
ordered = FALSE)
print(levels(zeno_pws_df$clean_sex))## [1] "Female" "Male" "Non-Binary"Drop healthy controls - no EDSS and T25FW
zeno_pws_df <- zeno_pws_df[grepl('MS', zeno_pws_df$demographic_diagnosis), ]
table(zeno_pws_df$demographic_diagnosis)##
## MS
## 217# convert infinite values to NA
zeno_pws_df[] <- lapply(zeno_pws_df, function(x) {
if (is.numeric(x)) replace(x, is.infinite(x), NA) else x
})# filter to unique IDs - one row per person, select first video
nrow(zeno_pws_df)## [1] 217zeno_pws_uniqueid_df <- zeno_pws_df %>%
arrange(visit_date_video) %>% # Sort by date
distinct(bw_id, .keep_all = TRUE) # Keep the first occurrence of each ID
nrow(zeno_pws_uniqueid_df)## [1] 152Missing Data
# count number of missing variables in preferred walking speed data frame
sum(is.na(zeno_pws_df$delta_pix_h_rel_median_pose_zv))## [1] 14sum(is.na(zeno_pws_df$stride_time_mean_sec_pose_zv))## [1] 48sum(is.na(zeno_pws_df$mean_cadence_step_per_min_pose_zv))## [1] 40sum(is.na(zeno_pws_df$stride_width_mean_cm_pose_zv))## [1] 40# stance and all double support measures
sum(is.na(zeno_pws_df$foot1_gait_cycle_time_mean_pose_zv))## [1] 159# ground truth preferred walk zeno mat data
sum(is.na(zeno_pws_df$PWS_cadencestepsminmean))## [1] 0sum(is.na(zeno_pws_df$bingoEHR_EDSS_measure_value))## [1] 0sum(is.na(zeno_pws_df$msfcEHR_T25FW.SPEED.AVG))## [1] 0# demographics
sum(is.na(zeno_pws_df$demoEHR_Age))## [1] 0sum(is.na(zeno_pws_df$demoEHR_DiseaseDuration)) ## [1] 0sum(is.na(zeno_pws_df$ms_dx_condensed))## [1] 0sum(is.na(zeno_pws_df$clean_ethnicity))## [1] 0sum(is.na(zeno_pws_df$clean_sex))## [1] 0sum(is.na(zeno_pws_df$ms_dx_condensed))## [1] 0Fast walking speed videos
zeno_fw_df <- read.csv("C:/Users/mmccu/Box/MM_Personal/5_Projects/BoveLab/3_Data_and_Code/gait_bw_zeno_home_analysis/004/000_merged_cleaned_data/zv_bw_merged_gait_vertical_FW_1_clean.csv")
table(zeno_fw_df$task_pose)##
## gait_vertical_FW_1
## 218# assign levels to categorical variables
table(zeno_fw_df$race_ethnicity_clean)##
## Asian Black Or African American Hispanic or Latino
## 17 14 20
## Other/Unknown/Declined White Not Hispanic
## 18 149zeno_fw_df$race_ethnicity_clean <- factor(zeno_fw_df$race_ethnicity_clean,
levels = c('Asian',
'Black Or African American',
'Hispanic or Latino',
'White Not Hispanic',
'Other/Unknown/Declined'),
ordered = FALSE)
print(levels(zeno_fw_df$race_ethnicity_clean))## [1] "Asian" "Black Or African American"
## [3] "Hispanic or Latino" "White Not Hispanic"
## [5] "Other/Unknown/Declined"table(zeno_fw_df$ms_dx_condensed)##
## MS, Subtype Not Specified Progressive MS RRMS
## 2 37 179zeno_fw_df$ms_dx_condensed <- factor(zeno_fw_df$ms_dx_condensed,
levels = c('RRMS',
'Progressive MS',
'MS, Subtype Not Specified'),
ordered = FALSE)
print(levels(zeno_fw_df$ms_dx_condensed))## [1] "RRMS" "Progressive MS"
## [3] "MS, Subtype Not Specified"table(zeno_fw_df$clean_sex)##
## Female Male Non-Binary
## 163 53 2zeno_fw_df$clean_sex <- factor(zeno_fw_df$clean_sex,
levels = c('Female',
'Male',
'Non-Binary'),
ordered = FALSE)
print(levels(zeno_fw_df$clean_sex))## [1] "Female" "Male" "Non-Binary"zeno_fw_df <- zeno_fw_df[grepl('MS', zeno_fw_df$demographic_diagnosis), ]
table(zeno_fw_df$demographic_diagnosis)##
## MS
## 218# convert infinite values to NA
zeno_fw_df[] <- lapply(zeno_fw_df, function(x) {
if (is.numeric(x)) replace(x, is.infinite(x), NA) else x
})# filter to unique IDs - one row per person, select first video
nrow(zeno_fw_df)## [1] 218zeno_fw_uniqueid_df <- zeno_fw_df %>%
arrange(visit_date_video) %>% # Sort by date
distinct(bw_id, .keep_all = TRUE) # Keep the first occurrence of each ID
nrow(zeno_fw_uniqueid_df)## [1] 152# count number of missing variables in fast walking speed data frame
sum(is.na(zeno_fw_df$delta_pix_h_rel_median_pose_zv))## [1] 3sum(is.na(zeno_fw_df$stride_time_mean_sec_pose_zv))## [1] 53sum(is.na(zeno_fw_df$mean_cadence_step_per_min_pose_zv))## [1] 46sum(is.na(zeno_fw_df$stride_width_mean_cm_pose_zv))## [1] 47# stance and all double support measures
sum(is.na(zeno_fw_df$foot1_gait_cycle_time_mean_pose_zv))## [1] 186# ground truth preferred walk zeno mat data
sum(is.na(zeno_fw_df$PWS_cadencestepsminmean))## [1] 0sum(is.na(zeno_fw_df$bingoEHR_EDSS_measure_value))## [1] 0sum(is.na(zeno_fw_df$msfcEHR_T25FW.SPEED.AVG))## [1] 0# demographics
sum(is.na(zeno_fw_df$demoEHR_Age))## [1] 0sum(is.na(zeno_fw_df$demoEHR_DiseaseDuration)) ## [1] 0sum(is.na(zeno_fw_df$race_ethnicity_clean))## [1] 0sum(is.na(zeno_fw_df$clean_sex))## [1] 0sum(is.na(zeno_fw_df$ms_dx_condensed))## [1] 0Load and format home video data
Some participants have videos from multiple timepoints (baseline and follow up) At each timepoint, participants sent back two videos - one turning to the right (gait_vertical_right task) and one turning to the left (gait_vertical_left task)
home_df <- read.csv("C:/Users/mmccu/Box/MM_Personal/5_Projects/BoveLab/3_Data_and_Code/gait_bw_zeno_home_analysis/004/000_merged_cleaned_data/hv_bw_merged_clean.csv")
nrow(home_df)## [1] 63table(home_df$demographic_diagnosis)##
## MS
## 63table(home_df$task_pose_hv)##
## gait_vertical_left gait_vertical_right
## 31 32# assign levels to categorical variables
table(home_df$race_ethnicity_clean)##
## Asian Black Or African American Hispanic or Latino
## 4 2 2
## Other/Unknown/Declined White Not Hispanic
## 9 46home_df$race_ethnicity_clean <- factor(home_df$race_ethnicity_clean,
levels = c('Asian',
'Black Or African American',
'Hispanic or Latino',
'White Not Hispanic',
'Other/Unknown/Declined'),
ordered = FALSE)
print(levels(home_df$race_ethnicity_clean))## [1] "Asian" "Black Or African American"
## [3] "Hispanic or Latino" "White Not Hispanic"
## [5] "Other/Unknown/Declined"table(home_df$ms_dx_condensed)##
## MS, Subtype Not Specified Progressive MS RRMS
## 4 7 52home_df$ms_dx_condensed <- factor(home_df$ms_dx_condensed,
levels = c('RRMS',
'Progressive MS',
'MS, Subtype Not Specified'),
ordered = FALSE)
print(levels(home_df$ms_dx_condensed))## [1] "RRMS" "Progressive MS"
## [3] "MS, Subtype Not Specified"table(home_df$clean_sex)##
## Female Male Non-Binary
## 52 9 2home_df$clean_sex <- factor(home_df$clean_sex,
levels = c('Female',
'Male',
'Non-Binary'),
ordered = FALSE)
print(levels(home_df$clean_sex))## [1] "Female" "Male" "Non-Binary"# convert infinite values to NA
home_df[] <- lapply(home_df, function(x) {
if (is.numeric(x)) replace(x, is.infinite(x), NA) else x
})Group by left and right turning videos, unique IDs
# right turning videos
home_r_df <- home_df[grepl('gait_vertical_right', home_df$task_pose_hv), ]
nrow(home_r_df)## [1] 32table(home_r_df$task_pose_hv)##
## gait_vertical_right
## 32# left turning videos
home_l_df <- home_df[grepl('gait_vertical_left', home_df$task_pose_hv), ]
nrow(home_l_df)## [1] 31table(home_l_df$task_pose_hv)##
## gait_vertical_left
## 31# filter to unique IDs - one row per person, select first video
home_uniqueid_df <- home_df %>%
arrange(visit_date_video) %>% # Sort by date
distinct(bw_id, .keep_all = TRUE) # Keep the first occurrence of each ID
nrow(home_uniqueid_df)## [1] 30# count number of missing variables in home video data frame
sum(is.na(home_df$delta_pix_h_rel_median_pose_hv))## [1] 4sum(is.na(home_df$stride_time_mean_sec_pose_hv))## [1] 13sum(is.na(home_df$mean_cadence_step_per_min_pose_hv))## [1] 12sum(is.na(home_df$stride_width_mean_cm_pose_hv))## [1] 12# stance and all double support measures
sum(is.na(home_df$foot1_gait_cycle_time_mean_pose_hv))## [1] 32# ground truth preferred walk zeno mat data
sum(is.na(home_df$PWS_cadencestepsminmean))## [1] 0sum(is.na(home_df$bingoEHR_EDSS_measure_value))## [1] 0sum(is.na(home_df$msfcEHR_T25FW.SPEED.AVG))## [1] 0# demographics
sum(is.na(home_df$demoEHR_Age))## [1] 0sum(is.na(home_df$demoEHR_DiseaseDuration)) ## [1] 0sum(is.na(home_df$race_ethnicity_clean))## [1] 0sum(is.na(home_df$clean_sex))## [1] 0sum(is.na(home_df$ms_dx_condensed))## [1] 0Function - Linear Regression Models
metric_regression <- function(data_frame, outcome, predictor) {
outcome_string <- deparse(substitute(outcome))
predictor_string <- deparse(substitute(predictor))
data_frame_string <- deparse(substitute(data_frame))
print(paste('Data Frame: ', data_frame_string))
# plot
print(ggplot(data = data_frame, aes_string(x = predictor_string,
y = outcome_string,
color = "demoEHR_Age",
shape = "ms_dx_condensed")) + geom_point())
# univariate model - just metric and outcome
outcome_predictor_string <- paste(outcome_string, "~", predictor_string, collapse = " ")
print(outcome_predictor_string)
model_1 <- lm(as.formula(outcome_predictor_string),
data = data_frame)
print(summary(model_1))
hist(resid(model_1))
# confounding: model + disease info and demographics
additional_vars <- "demoEHR_Age + demoEHR_DiseaseDuration +
ms_dx_condensed +
race_ethnicity_clean +
clean_sex"
full_formula_string = paste(outcome_predictor_string, "+", additional_vars, collapse = " ")
print(full_formula_string)
model_2 <- lm(as.formula(full_formula_string),
data = data_frame)
print(summary(model_2))
hist(resid(model_2))
}Linear Regression - T25FW
Transform T25FW
PWS, FW, and Home Videos
# preferred walking speed videos
zeno_pws_df$t25fw_log <- log(zeno_pws_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = zeno_pws_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = zeno_pws_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# unique IDS
zeno_pws_uniqueid_df$t25fw_log <- log(zeno_pws_uniqueid_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = zeno_pws_uniqueid_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = zeno_pws_uniqueid_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# fast walking speed videos
zeno_fw_df$t25fw_log <- log(zeno_fw_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = zeno_fw_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = zeno_fw_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# unique IDs
zeno_fw_uniqueid_df$t25fw_log <- log(zeno_fw_uniqueid_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = zeno_fw_uniqueid_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = zeno_fw_uniqueid_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# home - all videos
home_df$t25fw_log <- log(home_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = home_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = home_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# home - right
home_r_df$t25fw_log <- log(home_r_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = home_r_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = home_r_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# home - left
home_l_df$t25fw_log <- log(home_l_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = home_l_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = home_l_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# home - unique
home_uniqueid_df$t25fw_log <- log(home_uniqueid_df$msfcEHR_T25FW.SPEED.AVG)
ggplot(data = home_uniqueid_df, mapping = aes(msfcEHR_T25FW.SPEED.AVG)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
ggplot(data = home_uniqueid_df, mapping = aes(t25fw_log)) +
geom_histogram()## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## Preferred Walking Speed –> T25FW
PWS Demographics and Disease Info –> Log T25FW
# Preferred walking speed participants
# model
pws_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
race_ethnicity_clean +
clean_sex,
data = zeno_pws_df)
summary(pws_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + race_ethnicity_clean + clean_sex, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.05263 -0.21089 -0.05063 0.13211 1.78009
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.343709 0.140298 9.578
## demoEHR_Age 0.007483 0.002850 2.626
## demoEHR_DiseaseDuration -0.004477 0.003750 -1.194
## ms_dx_condensedProgressive MS 0.468407 0.077454 6.048
## ms_dx_condensedMS, Subtype Not Specified -0.029325 0.278886 -0.105
## race_ethnicity_cleanBlack Or African American 0.182299 0.142230 1.282
## race_ethnicity_cleanHispanic or Latino -0.060882 0.128107 -0.475
## race_ethnicity_cleanWhite Not Hispanic -0.159167 0.104390 -1.525
## race_ethnicity_cleanOther/Unknown/Declined -0.087477 0.131936 -0.663
## clean_sexMale -0.050412 0.063415 -0.795
## clean_sexNon-Binary 0.007581 0.276860 0.027
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## demoEHR_Age 0.00929 **
## demoEHR_DiseaseDuration 0.23383
## ms_dx_condensedProgressive MS 6.82e-09 ***
## ms_dx_condensedMS, Subtype Not Specified 0.91636
## race_ethnicity_cleanBlack Or African American 0.20138
## race_ethnicity_cleanHispanic or Latino 0.63512
## race_ethnicity_cleanWhite Not Hispanic 0.12886
## race_ethnicity_cleanOther/Unknown/Declined 0.50805
## clean_sexMale 0.42756
## clean_sexNon-Binary 0.97818
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3848 on 206 degrees of freedom
## Multiple R-squared: 0.2792, Adjusted R-squared: 0.2442
## F-statistic: 7.98 on 10 and 206 DF, p-value: 7.833e-11hist(resid(pws_dem_model))
# model - unique IDS
pws_dem_model_2 <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
race_ethnicity_clean +
clean_sex,
data = zeno_pws_uniqueid_df)
summary(pws_dem_model_2)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + race_ethnicity_clean + clean_sex, data = zeno_pws_uniqueid_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.92748 -0.18385 -0.04131 0.10597 1.55919
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.404395 0.163024 8.615
## demoEHR_Age 0.006096 0.003078 1.980
## demoEHR_DiseaseDuration -0.002272 0.004540 -0.500
## ms_dx_condensedProgressive MS 0.491164 0.086659 5.668
## ms_dx_condensedMS, Subtype Not Specified -0.033748 0.273835 -0.123
## race_ethnicity_cleanBlack Or African American 0.054728 0.162035 0.338
## race_ethnicity_cleanHispanic or Latino -0.036624 0.149956 -0.244
## race_ethnicity_cleanWhite Not Hispanic -0.175147 0.123557 -1.418
## race_ethnicity_cleanOther/Unknown/Declined -0.078325 0.158644 -0.494
## clean_sexMale -0.023110 0.072929 -0.317
## clean_sexNon-Binary -0.005586 0.380029 -0.015
## Pr(>|t|)
## (Intercept) 1.27e-14 ***
## demoEHR_Age 0.0496 *
## demoEHR_DiseaseDuration 0.6176
## ms_dx_condensedProgressive MS 7.86e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.9021
## race_ethnicity_cleanBlack Or African American 0.7361
## race_ethnicity_cleanHispanic or Latino 0.8074
## race_ethnicity_cleanWhite Not Hispanic 0.1585
## race_ethnicity_cleanOther/Unknown/Declined 0.6223
## clean_sexMale 0.7518
## clean_sexNon-Binary 0.9883
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3752 on 141 degrees of freedom
## Multiple R-squared: 0.2885, Adjusted R-squared: 0.238
## F-statistic: 5.717 on 10 and 141 DF, p-value: 3.586e-07hist(resid(pws_dem_model_2))
Univariate - PWS, delta pixel proxy –> T25fw
# Preferred Walking Speed
sum(is.finite(zeno_pws_df$delta_pix_h_rel_median_pose_zv))## [1] 203zeno_pws_df$log_delta_pix_h_rel_median_pose_zv <- log(zeno_pws_df$delta_pix_h_rel_median_pose_zv)
zeno_pws_df$sqrt_delta_pix_h_rel_median_pose_zv <- sqrt(zeno_pws_df$delta_pix_h_rel_median_pose_zv)
# log velproxy and log t25Fw
metric_regression(zeno_pws_df, t25fw_log, log_delta_pix_h_rel_median_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: `aes_string()` was deprecated in ggplot2 3.0.0.
## i Please use tidy evaluation idioms with `aes()`.
## i See also `vignette("ggplot2-in-packages")` for more information.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## Warning: Removed 14 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.17314 -0.20739 -0.02811 0.17522 1.48275
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.90555 0.07602 11.91 <2e-16 ***
## log_delta_pix_h_rel_median_pose_zv -0.51691 0.05066 -10.20 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3689 on 201 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.3412, Adjusted R-squared: 0.338
## F-statistic: 104.1 on 1 and 201 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.94418 -0.17936 -0.02638 0.15895 1.22540
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.651281 0.149829 4.347
## log_delta_pix_h_rel_median_pose_zv -0.416540 0.051460 -8.094
## demoEHR_Age 0.008772 0.002503 3.505
## demoEHR_DiseaseDuration -0.003550 0.003309 -1.073
## ms_dx_condensedProgressive MS 0.256603 0.076814 3.341
## ms_dx_condensedMS, Subtype Not Specified 0.022346 0.242698 0.092
## race_ethnicity_cleanBlack Or African American 0.254830 0.127236 2.003
## race_ethnicity_cleanHispanic or Latino 0.058290 0.118137 0.493
## race_ethnicity_cleanWhite Not Hispanic -0.096629 0.095827 -1.008
## race_ethnicity_cleanOther/Unknown/Declined -0.015309 0.120027 -0.128
## clean_sexMale -0.024896 0.057579 -0.432
## clean_sexNon-Binary 0.120376 0.338105 0.356
## Pr(>|t|)
## (Intercept) 2.24e-05 ***
## log_delta_pix_h_rel_median_pose_zv 6.59e-14 ***
## demoEHR_Age 0.00057 ***
## demoEHR_DiseaseDuration 0.28462
## ms_dx_condensedProgressive MS 0.00101 **
## ms_dx_condensedMS, Subtype Not Specified 0.92673
## race_ethnicity_cleanBlack Or African American 0.04661 *
## race_ethnicity_cleanHispanic or Latino 0.62229
## race_ethnicity_cleanWhite Not Hispanic 0.31455
## race_ethnicity_cleanOther/Unknown/Declined 0.89864
## clean_sexMale 0.66596
## clean_sexNon-Binary 0.72221
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3346 on 191 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.4851, Adjusted R-squared: 0.4554
## F-statistic: 16.36 on 11 and 191 DF, p-value: < 2.2e-16
# Benchmark - True PWS mat velocity and log25fw
metric_regression(zeno_pws_df, t25fw_log, PWS_velocitycmsecmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ PWS_velocitycmsecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.68111 -0.18501 0.00107 0.17512 1.27235
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.7684605 0.0723392 38.27 <2e-16 ***
## PWS_velocitycmsecmean -0.0109993 0.0006715 -16.38 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2959 on 215 degrees of freedom
## Multiple R-squared: 0.5551, Adjusted R-squared: 0.5531
## F-statistic: 268.3 on 1 and 215 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ PWS_velocitycmsecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.59572 -0.15358 -0.01625 0.16525 1.16504
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.5211704 0.1361926 18.512
## PWS_velocitycmsecmean -0.0097653 0.0007369 -13.252
## demoEHR_Age 0.0064232 0.0020981 3.061
## demoEHR_DiseaseDuration -0.0041353 0.0027587 -1.499
## ms_dx_condensedProgressive MS 0.1403804 0.0621267 2.260
## ms_dx_condensedMS, Subtype Not Specified 0.1097673 0.2054417 0.534
## race_ethnicity_cleanBlack Or African American -0.0781671 0.1064672 -0.734
## race_ethnicity_cleanHispanic or Latino -0.1197515 0.0943516 -1.269
## race_ethnicity_cleanWhite Not Hispanic -0.2099054 0.0768939 -2.730
## race_ethnicity_cleanOther/Unknown/Declined -0.1491848 0.0971753 -1.535
## clean_sexMale -0.0574213 0.0466571 -1.231
## clean_sexNon-Binary 0.0960789 0.2037927 0.471
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## PWS_velocitycmsecmean < 2e-16 ***
## demoEHR_Age 0.00250 **
## demoEHR_DiseaseDuration 0.13542
## ms_dx_condensedProgressive MS 0.02490 *
## ms_dx_condensedMS, Subtype Not Specified 0.59371
## race_ethnicity_cleanBlack Or African American 0.46367
## race_ethnicity_cleanHispanic or Latino 0.20581
## race_ethnicity_cleanWhite Not Hispanic 0.00689 **
## race_ethnicity_cleanOther/Unknown/Declined 0.12627
## clean_sexMale 0.21984
## clean_sexNon-Binary 0.63782
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2831 on 205 degrees of freedom
## Multiple R-squared: 0.6118, Adjusted R-squared: 0.591
## F-statistic: 29.37 on 11 and 205 DF, p-value: < 2.2e-16
# unique IDs
zeno_pws_uniqueid_df$log_delta_pix_h_rel_median_pose_zv <- log(zeno_pws_uniqueid_df$delta_pix_h_rel_median_pose_zv)
metric_regression(zeno_pws_uniqueid_df, t25fw_log, log_delta_pix_h_rel_median_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.95970 -0.18760 -0.04187 0.16277 1.39322
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.86851 0.08069 10.76 <2e-16 ***
## log_delta_pix_h_rel_median_pose_zv -0.56702 0.05558 -10.20 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3313 on 150 degrees of freedom
## Multiple R-squared: 0.4096, Adjusted R-squared: 0.4057
## F-statistic: 104.1 on 1 and 150 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.77069 -0.16971 -0.03483 0.12025 1.16477
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.752857 0.156198 4.820
## log_delta_pix_h_rel_median_pose_zv -0.492004 0.060006 -8.199
## demoEHR_Age 0.007514 0.002545 2.952
## demoEHR_DiseaseDuration -0.004691 0.003757 -1.249
## ms_dx_condensedProgressive MS 0.180178 0.080922 2.227
## ms_dx_condensedMS, Subtype Not Specified 0.036349 0.226041 0.161
## race_ethnicity_cleanBlack Or African American 0.075771 0.133683 0.567
## race_ethnicity_cleanHispanic or Latino -0.019307 0.123712 -0.156
## race_ethnicity_cleanWhite Not Hispanic -0.200893 0.101967 -1.970
## race_ethnicity_cleanOther/Unknown/Declined -0.168479 0.131322 -1.283
## clean_sexMale -0.003996 0.060202 -0.066
## clean_sexNon-Binary 0.103824 0.313759 0.331
## Pr(>|t|)
## (Intercept) 3.69e-06 ***
## log_delta_pix_h_rel_median_pose_zv 1.39e-13 ***
## demoEHR_Age 0.0037 **
## demoEHR_DiseaseDuration 0.2139
## ms_dx_condensedProgressive MS 0.0276 *
## ms_dx_condensedMS, Subtype Not Specified 0.8725
## race_ethnicity_cleanBlack Or African American 0.5718
## race_ethnicity_cleanHispanic or Latino 0.8762
## race_ethnicity_cleanWhite Not Hispanic 0.0508 .
## race_ethnicity_cleanOther/Unknown/Declined 0.2016
## clean_sexMale 0.9472
## clean_sexNon-Binary 0.7412
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3095 on 140 degrees of freedom
## Multiple R-squared: 0.5193, Adjusted R-squared: 0.4815
## F-statistic: 13.75 on 11 and 140 DF, p-value: < 2.2e-16
Univariate - PWS, stride time –> T25fw
# Preferred Walking speed
sum(is.finite(zeno_pws_df$stride_time_mean_sec_pose_zv))## [1] 169# stride time mean
ggplot(data = zeno_pws_df, aes(x = stride_time_mean_sec_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 48 rows containing missing values (`geom_point()`).
# stride time median
ggplot(data = zeno_pws_df, aes(x = stride_time_median_sec_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 48 rows containing missing values (`geom_point()`).
# stride time max
ggplot(data = zeno_pws_df, aes(x = stride_time_max_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 48 rows containing missing values (`geom_point()`).
# stride time min
ggplot(data = zeno_pws_df, aes(x = stride_time_min_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 48 rows containing missing values (`geom_point()`).
# stride time std
ggplot(data = zeno_pws_df, aes(x = stride_time_std_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 53 rows containing missing values (`geom_point()`).
# stride time cv
ggplot(data = zeno_pws_df, aes(x = stride_time_cv_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 53 rows containing missing values (`geom_point()`).
# Selected median, seemed most linear
metric_regression(zeno_pws_df, t25fw_log, stride_time_median_sec_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 48 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.76611 -0.20721 -0.07297 0.16050 1.24166
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.5683 0.1934 2.939 0.00376 **
## stride_time_median_sec_pose_zv 0.9197 0.1707 5.388 2.39e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3274 on 167 degrees of freedom
## (48 observations deleted due to missingness)
## Multiple R-squared: 0.1481, Adjusted R-squared: 0.143
## F-statistic: 29.03 on 1 and 167 DF, p-value: 2.39e-07
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.59735 -0.19259 -0.06107 0.12804 1.20275
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.546523 0.225362 2.425
## stride_time_median_sec_pose_zv 0.781865 0.162184 4.821
## demoEHR_Age 0.005610 0.002506 2.238
## demoEHR_DiseaseDuration 0.003138 0.003365 0.933
## ms_dx_condensedProgressive MS 0.193693 0.073630 2.631
## ms_dx_condensedMS, Subtype Not Specified -0.036535 0.213422 -0.171
## race_ethnicity_cleanBlack Or African American 0.248300 0.122594 2.025
## race_ethnicity_cleanHispanic or Latino -0.187214 0.111049 -1.686
## race_ethnicity_cleanWhite Not Hispanic -0.196061 0.093206 -2.104
## race_ethnicity_cleanOther/Unknown/Declined -0.081488 0.130867 -0.623
## clean_sexMale -0.057530 0.053015 -1.085
## clean_sexNon-Binary 0.048819 0.296606 0.165
## Pr(>|t|)
## (Intercept) 0.01644 *
## stride_time_median_sec_pose_zv 3.35e-06 ***
## demoEHR_Age 0.02660 *
## demoEHR_DiseaseDuration 0.35247
## ms_dx_condensedProgressive MS 0.00937 **
## ms_dx_condensedMS, Subtype Not Specified 0.86430
## race_ethnicity_cleanBlack Or African American 0.04452 *
## race_ethnicity_cleanHispanic or Latino 0.09381 .
## race_ethnicity_cleanWhite Not Hispanic 0.03701 *
## race_ethnicity_cleanOther/Unknown/Declined 0.53440
## clean_sexMale 0.27951
## clean_sexNon-Binary 0.86948
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2933 on 157 degrees of freedom
## (48 observations deleted due to missingness)
## Multiple R-squared: 0.3574, Adjusted R-squared: 0.3124
## F-statistic: 7.939 on 11 and 157 DF, p-value: 6.801e-11
# benchmark - true PWS stride time
metric_regression(zeno_pws_df, t25fw_log, PWS_stridetimesecmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ PWS_stridetimesecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.72699 -0.18864 -0.01653 0.13678 1.39715
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.44946 0.07601 5.913 1.31e-08 ***
## PWS_stridetimesecmean 0.96084 0.05961 16.117 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2985 on 215 degrees of freedom
## Multiple R-squared: 0.5472, Adjusted R-squared: 0.545
## F-statistic: 259.8 on 1 and 215 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ PWS_stridetimesecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.59456 -0.17436 -0.03791 0.13411 1.36432
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.367559 0.119960 3.064
## PWS_stridetimesecmean 0.877540 0.060819 14.429
## demoEHR_Age 0.006747 0.002013 3.352
## demoEHR_DiseaseDuration 0.002570 0.002692 0.955
## ms_dx_condensedProgressive MS 0.136581 0.059328 2.302
## ms_dx_condensedMS, Subtype Not Specified -0.009244 0.196923 -0.047
## race_ethnicity_cleanBlack Or African American 0.029873 0.100981 0.296
## race_ethnicity_cleanHispanic or Latino -0.207634 0.091025 -2.281
## race_ethnicity_cleanWhite Not Hispanic -0.237048 0.073906 -3.207
## race_ethnicity_cleanOther/Unknown/Declined -0.144949 0.093243 -1.555
## clean_sexMale -0.038496 0.044785 -0.860
## clean_sexNon-Binary 0.126224 0.195661 0.645
## Pr(>|t|)
## (Intercept) 0.002477 **
## PWS_stridetimesecmean < 2e-16 ***
## demoEHR_Age 0.000955 ***
## demoEHR_DiseaseDuration 0.340867
## ms_dx_condensedProgressive MS 0.022332 *
## ms_dx_condensedMS, Subtype Not Specified 0.962605
## race_ethnicity_cleanBlack Or African American 0.767659
## race_ethnicity_cleanHispanic or Latino 0.023572 *
## race_ethnicity_cleanWhite Not Hispanic 0.001554 **
## race_ethnicity_cleanOther/Unknown/Declined 0.121602
## clean_sexMale 0.391026
## clean_sexNon-Binary 0.519573
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2717 on 205 degrees of freedom
## Multiple R-squared: 0.6424, Adjusted R-squared: 0.6232
## F-statistic: 33.48 on 11 and 205 DF, p-value: < 2.2e-16
# video unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, stride_time_median_sec_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 37 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.48517 -0.23439 -0.05909 0.14529 1.13224
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.7273 0.2230 3.262 0.001465 **
## stride_time_median_sec_pose_zv 0.7893 0.1968 4.011 0.000109 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3269 on 113 degrees of freedom
## (37 observations deleted due to missingness)
## Multiple R-squared: 0.1246, Adjusted R-squared: 0.1169
## F-statistic: 16.09 on 1 and 113 DF, p-value: 0.0001089
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.49847 -0.19754 -0.06398 0.15701 1.07644
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.879402 0.280378 3.136
## stride_time_median_sec_pose_zv 0.582500 0.188486 3.090
## demoEHR_Age 0.004579 0.002834 1.616
## demoEHR_DiseaseDuration 0.003125 0.004120 0.758
## ms_dx_condensedProgressive MS 0.263836 0.087106 3.029
## ms_dx_condensedMS, Subtype Not Specified -0.032504 0.218140 -0.149
## race_ethnicity_cleanBlack Or African American 0.126946 0.157660 0.805
## race_ethnicity_cleanHispanic or Latino -0.226582 0.149075 -1.520
## race_ethnicity_cleanWhite Not Hispanic -0.251659 0.128450 -1.959
## race_ethnicity_cleanOther/Unknown/Declined -0.183476 0.167339 -1.096
## clean_sexMale -0.055093 0.064653 -0.852
## clean_sexNon-Binary 0.034862 0.301936 0.115
## Pr(>|t|)
## (Intercept) 0.00223 **
## stride_time_median_sec_pose_zv 0.00257 **
## demoEHR_Age 0.10921
## demoEHR_DiseaseDuration 0.44998
## ms_dx_condensedProgressive MS 0.00310 **
## ms_dx_condensedMS, Subtype Not Specified 0.88184
## race_ethnicity_cleanBlack Or African American 0.42257
## race_ethnicity_cleanHispanic or Latino 0.13159
## race_ethnicity_cleanWhite Not Hispanic 0.05279 .
## race_ethnicity_cleanOther/Unknown/Declined 0.27545
## clean_sexMale 0.39612
## clean_sexNon-Binary 0.90830
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2972 on 103 degrees of freedom
## (37 observations deleted due to missingness)
## Multiple R-squared: 0.3405, Adjusted R-squared: 0.2701
## F-statistic: 4.835 on 11 and 103 DF, p-value: 5.552e-06
Univariate - PWS, Cadence –> T25fw
# PWS
sum(is.finite(zeno_pws_df$mean_cadence_step_per_min_pose_zv))## [1] 177# cadence model
metric_regression(zeno_pws_df, t25fw_log, mean_cadence_step_per_min_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 40 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.79691 -0.22803 -0.05002 0.13301 1.85437
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.628380 0.194369 13.523 < 2e-16 ***
## mean_cadence_step_per_min_pose_zv -0.009757 0.001859 -5.249 4.39e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3677 on 175 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.136, Adjusted R-squared: 0.1311
## F-statistic: 27.55 on 1 and 175 DF, p-value: 4.388e-07
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.58271 -0.18362 -0.03909 0.13333 1.42880
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.296e+00 2.286e-01 10.044
## mean_cadence_step_per_min_pose_zv -8.501e-03 1.740e-03 -4.886
## demoEHR_Age 7.304e-03 2.658e-03 2.748
## demoEHR_DiseaseDuration -3.152e-05 3.539e-03 -0.009
## ms_dx_condensedProgressive MS 2.910e-01 7.578e-02 3.840
## ms_dx_condensedMS, Subtype Not Specified 2.094e-02 2.327e-01 0.090
## race_ethnicity_cleanBlack Or African American 2.452e-01 1.332e-01 1.841
## race_ethnicity_cleanHispanic or Latino -1.018e-01 1.197e-01 -0.850
## race_ethnicity_cleanWhite Not Hispanic -2.432e-01 1.013e-01 -2.402
## race_ethnicity_cleanOther/Unknown/Declined -1.798e-01 1.411e-01 -1.274
## clean_sexMale -1.155e-01 5.709e-02 -2.023
## clean_sexNon-Binary 7.016e-02 3.233e-01 0.217
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## mean_cadence_step_per_min_pose_zv 2.42e-06 ***
## demoEHR_Age 0.006658 **
## demoEHR_DiseaseDuration 0.992904
## ms_dx_condensedProgressive MS 0.000175 ***
## ms_dx_condensedMS, Subtype Not Specified 0.928390
## race_ethnicity_cleanBlack Or African American 0.067365 .
## race_ethnicity_cleanHispanic or Latino 0.396462
## race_ethnicity_cleanWhite Not Hispanic 0.017419 *
## race_ethnicity_cleanOther/Unknown/Declined 0.204411
## clean_sexMale 0.044684 *
## clean_sexNon-Binary 0.828475
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3196 on 165 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.3846, Adjusted R-squared: 0.3436
## F-statistic: 9.375 on 11 and 165 DF, p-value: 5.554e-13
# ground truth cadence from mat
metric_regression(zeno_pws_df, t25fw_log, PWS_cadencestepsminmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ PWS_cadencestepsminmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.78923 -0.19583 -0.02581 0.18096 1.27164
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.482224 0.138322 25.18 <2e-16 ***
## PWS_cadencestepsminmean -0.018329 0.001351 -13.56 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3256 on 215 degrees of freedom
## Multiple R-squared: 0.4611, Adjusted R-squared: 0.4586
## F-statistic: 184 on 1 and 215 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ PWS_cadencestepsminmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.5824 -0.1870 -0.0275 0.1530 1.2647
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 3.108222 0.176990 17.562
## PWS_cadencestepsminmean -0.016898 0.001357 -12.456
## demoEHR_Age 0.008182 0.002156 3.795
## demoEHR_DiseaseDuration 0.002822 0.002896 0.975
## ms_dx_condensedProgressive MS 0.162101 0.063530 2.552
## ms_dx_condensedMS, Subtype Not Specified 0.029441 0.210971 0.140
## race_ethnicity_cleanBlack Or African American -0.028201 0.108887 -0.259
## race_ethnicity_cleanHispanic or Latino -0.157025 0.097193 -1.616
## race_ethnicity_cleanWhite Not Hispanic -0.264972 0.079405 -3.337
## race_ethnicity_cleanOther/Unknown/Declined -0.164333 0.099972 -1.644
## clean_sexMale -0.094579 0.048091 -1.967
## clean_sexNon-Binary 0.144435 0.209675 0.689
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## PWS_cadencestepsminmean < 2e-16 ***
## demoEHR_Age 0.000194 ***
## demoEHR_DiseaseDuration 0.330926
## ms_dx_condensedProgressive MS 0.011453 *
## ms_dx_condensedMS, Subtype Not Specified 0.889154
## race_ethnicity_cleanBlack Or African American 0.795903
## race_ethnicity_cleanHispanic or Latino 0.107720
## race_ethnicity_cleanWhite Not Hispanic 0.001006 **
## race_ethnicity_cleanOther/Unknown/Declined 0.101752
## clean_sexMale 0.050573 .
## clean_sexNon-Binary 0.491696
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.291 on 205 degrees of freedom
## Multiple R-squared: 0.5897, Adjusted R-squared: 0.5677
## F-statistic: 26.79 on 11 and 205 DF, p-value: < 2.2e-16
# unique ID
metric_regression(zeno_pws_uniqueid_df, t25fw_log, mean_cadence_step_per_min_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 29 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.65924 -0.24384 -0.07236 0.11962 1.88288
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.455456 0.234817 10.457 < 2e-16 ***
## mean_cadence_step_per_min_pose_zv -0.007971 0.002262 -3.524 0.000601 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3879 on 121 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.09307, Adjusted R-squared: 0.08558
## F-statistic: 12.42 on 1 and 121 DF, p-value: 0.0006012
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.64415 -0.22443 -0.03427 0.11734 1.36220
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.1619560 0.2881680 7.502
## mean_cadence_step_per_min_pose_zv -0.0063258 0.0020926 -3.023
## demoEHR_Age 0.0069087 0.0030653 2.254
## demoEHR_DiseaseDuration -0.0005588 0.0044213 -0.126
## ms_dx_condensedProgressive MS 0.3840172 0.0911174 4.215
## ms_dx_condensedMS, Subtype Not Specified 0.0249596 0.2455790 0.102
## race_ethnicity_cleanBlack Or African American 0.1123807 0.1773471 0.634
## race_ethnicity_cleanHispanic or Latino -0.1382831 0.1652262 -0.837
## race_ethnicity_cleanWhite Not Hispanic -0.3253947 0.1441959 -2.257
## race_ethnicity_cleanOther/Unknown/Declined -0.2666967 0.1862701 -1.432
## clean_sexMale -0.1067168 0.0713328 -1.496
## clean_sexNon-Binary 0.0603212 0.3398157 0.178
## Pr(>|t|)
## (Intercept) 1.66e-11 ***
## mean_cadence_step_per_min_pose_zv 0.00311 **
## demoEHR_Age 0.02617 *
## demoEHR_DiseaseDuration 0.89965
## ms_dx_condensedProgressive MS 5.12e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.91923
## race_ethnicity_cleanBlack Or African American 0.52760
## race_ethnicity_cleanHispanic or Latino 0.40443
## race_ethnicity_cleanWhite Not Hispanic 0.02599 *
## race_ethnicity_cleanOther/Unknown/Declined 0.15502
## clean_sexMale 0.13748
## clean_sexNon-Binary 0.85943
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3343 on 111 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3819, Adjusted R-squared: 0.3207
## F-statistic: 6.235 on 11 and 111 DF, p-value: 6.376e-08
Univariate - PWS, Stride Width –> T25fw
sum(is.finite(zeno_pws_df$stride_width_mean_cm_pose_zv))## [1] 177# Preferred Walking Speed
ggplot(data = zeno_pws_df, aes(x = stride_width_mean_cm_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_median_cm_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_min_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_max_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_std_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 42 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_cv_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 42 rows containing missing values (`geom_point()`).
## model
metric_regression(zeno_pws_df, t25fw_log, stride_width_median_cm_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 40 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.69520 -0.22012 -0.06513 0.11031 2.28879
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.183267 0.107949 10.961 < 2e-16 ***
## stride_width_median_cm_pose_zv 0.034898 0.008352 4.178 4.63e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3772 on 175 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.09072, Adjusted R-squared: 0.08552
## F-statistic: 17.46 on 1 and 175 DF, p-value: 4.627e-05
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.66382 -0.20045 -0.05566 0.15620 1.71245
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.135503 0.175597 6.467
## stride_width_median_cm_pose_zv 0.019155 0.008008 2.392
## demoEHR_Age 0.007323 0.002796 2.620
## demoEHR_DiseaseDuration -0.002412 0.003688 -0.654
## ms_dx_condensedProgressive MS 0.337942 0.079970 4.226
## ms_dx_condensedMS, Subtype Not Specified -0.060155 0.245001 -0.246
## race_ethnicity_cleanBlack Or African American 0.271156 0.139950 1.938
## race_ethnicity_cleanHispanic or Latino -0.053535 0.127092 -0.421
## race_ethnicity_cleanWhite Not Hispanic -0.188346 0.107019 -1.760
## race_ethnicity_cleanOther/Unknown/Declined -0.011352 0.145930 -0.078
## clean_sexMale -0.097167 0.059913 -1.622
## clean_sexNon-Binary -0.028108 0.339871 -0.083
## Pr(>|t|)
## (Intercept) 1.09e-09 ***
## stride_width_median_cm_pose_zv 0.01789 *
## demoEHR_Age 0.00962 **
## demoEHR_DiseaseDuration 0.51396
## ms_dx_condensedProgressive MS 3.93e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.80635
## race_ethnicity_cleanBlack Or African American 0.05439 .
## race_ethnicity_cleanHispanic or Latino 0.67413
## race_ethnicity_cleanWhite Not Hispanic 0.08027 .
## race_ethnicity_cleanOther/Unknown/Declined 0.93809
## clean_sexMale 0.10675
## clean_sexNon-Binary 0.93419
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3361 on 165 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.3192, Adjusted R-squared: 0.2738
## F-statistic: 7.032 on 11 and 165 DF, p-value: 1.033e-09
## pressure mat stride width model
metric_regression(zeno_pws_df, t25fw_log, PWS_stridewidthcmmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ PWS_stridewidthcmmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.61811 -0.22611 -0.09524 0.12265 2.08325
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.188185 0.072603 16.365 < 2e-16 ***
## PWS_stridewidthcmmean 0.045545 0.006925 6.577 3.59e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4048 on 215 degrees of freedom
## Multiple R-squared: 0.1675, Adjusted R-squared: 0.1636
## F-statistic: 43.26 on 1 and 215 DF, p-value: 3.592e-10
## [1] "t25fw_log ~ PWS_stridewidthcmmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.85936 -0.19830 -0.04167 0.12074 1.66149
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.943975 0.152398 6.194
## PWS_stridewidthcmmean 0.035351 0.006730 5.253
## demoEHR_Age 0.007900 0.002683 2.944
## demoEHR_DiseaseDuration -0.005228 0.003532 -1.480
## ms_dx_condensedProgressive MS 0.382962 0.074686 5.128
## ms_dx_condensedMS, Subtype Not Specified 0.119986 0.263996 0.454
## race_ethnicity_cleanBlack Or African American 0.129692 0.134228 0.966
## race_ethnicity_cleanHispanic or Latino -0.010466 0.120944 -0.087
## race_ethnicity_cleanWhite Not Hispanic -0.076788 0.099486 -0.772
## race_ethnicity_cleanOther/Unknown/Declined -0.008564 0.125071 -0.068
## clean_sexMale -0.082657 0.059996 -1.378
## clean_sexNon-Binary 0.036218 0.260612 0.139
## Pr(>|t|)
## (Intercept) 3.15e-09 ***
## PWS_stridewidthcmmean 3.75e-07 ***
## demoEHR_Age 0.00361 **
## demoEHR_DiseaseDuration 0.14035
## ms_dx_condensedProgressive MS 6.77e-07 ***
## ms_dx_condensedMS, Subtype Not Specified 0.64995
## race_ethnicity_cleanBlack Or African American 0.33508
## race_ethnicity_cleanHispanic or Latino 0.93112
## race_ethnicity_cleanWhite Not Hispanic 0.44110
## race_ethnicity_cleanOther/Unknown/Declined 0.94547
## clean_sexMale 0.16979
## clean_sexNon-Binary 0.88961
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3621 on 205 degrees of freedom
## Multiple R-squared: 0.3647, Adjusted R-squared: 0.3306
## F-statistic: 10.7 on 11 and 205 DF, p-value: 1.767e-15
# unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, stride_width_median_cm_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 29 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.57389 -0.25335 -0.08085 0.09591 2.19408
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.35481 0.13211 10.255 <2e-16 ***
## stride_width_median_cm_pose_zv 0.02281 0.01027 2.222 0.0281 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3992 on 121 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.0392, Adjusted R-squared: 0.03126
## F-statistic: 4.937 on 1 and 121 DF, p-value: 0.02814
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.79622 -0.20573 -0.05697 0.12050 1.47791
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.445732 0.225611 6.408
## stride_width_median_cm_pose_zv 0.003154 0.009727 0.324
## demoEHR_Age 0.006638 0.003186 2.083
## demoEHR_DiseaseDuration -0.001506 0.004586 -0.328
## ms_dx_condensedProgressive MS 0.453504 0.095801 4.734
## ms_dx_condensedMS, Subtype Not Specified -0.020583 0.255636 -0.081
## race_ethnicity_cleanBlack Or African American 0.133200 0.184329 0.723
## race_ethnicity_cleanHispanic or Latino -0.110095 0.173265 -0.635
## race_ethnicity_cleanWhite Not Hispanic -0.290387 0.150030 -1.936
## race_ethnicity_cleanOther/Unknown/Declined -0.153722 0.190282 -0.808
## clean_sexMale -0.089729 0.074282 -1.208
## clean_sexNon-Binary 0.002518 0.353160 0.007
## Pr(>|t|)
## (Intercept) 3.68e-09 ***
## stride_width_median_cm_pose_zv 0.7464
## demoEHR_Age 0.0395 *
## demoEHR_DiseaseDuration 0.7432
## ms_dx_condensedProgressive MS 6.55e-06 ***
## ms_dx_condensedMS, Subtype Not Specified 0.9360
## race_ethnicity_cleanBlack Or African American 0.4714
## race_ethnicity_cleanHispanic or Latino 0.5265
## race_ethnicity_cleanWhite Not Hispanic 0.0555 .
## race_ethnicity_cleanOther/Unknown/Declined 0.4209
## clean_sexMale 0.2296
## clean_sexNon-Binary 0.9943
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3477 on 111 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3317, Adjusted R-squared: 0.2654
## F-statistic: 5.008 on 11 and 111 DF, p-value: 2.677e-06
Univariate - PWS, Stance Time –> T25fw
Stance/swing/double/single support measures not calculated for all participants
# PWS
sum(is.finite(zeno_pws_df$foot1_stance_time_mean_pose_zv))## [1] 58ggplot(data = zeno_pws_df, aes(x = foot1_stance_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_stance_per_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
#video model
metric_regression(zeno_pws_df, t25fw_log, foot1_stance_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 159 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.5121 -0.1483 -0.0231 0.1114 0.9443
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.0634 0.2105 5.052 4.97e-06 ***
## foot1_stance_time_mean_pose_zv 0.7250 0.2745 2.641 0.0107 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2743 on 56 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.1107, Adjusted R-squared: 0.09487
## F-statistic: 6.974 on 1 and 56 DF, p-value: 0.0107
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.50093 -0.15215 0.00611 0.13374 0.80566
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.017779 0.272871 3.730
## foot1_stance_time_mean_pose_zv 0.625083 0.273156 2.288
## demoEHR_Age 0.002425 0.004031 0.602
## demoEHR_DiseaseDuration 0.003317 0.005412 0.613
## ms_dx_condensedProgressive MS 0.181272 0.130976 1.384
## race_ethnicity_cleanBlack Or African American 0.411453 0.188757 2.180
## race_ethnicity_cleanHispanic or Latino -0.089447 0.137070 -0.653
## race_ethnicity_cleanWhite Not Hispanic -0.085178 0.127121 -0.670
## clean_sexMale 0.002421 0.084688 0.029
## Pr(>|t|)
## (Intercept) 0.000498 ***
## foot1_stance_time_mean_pose_zv 0.026467 *
## demoEHR_Age 0.550174
## demoEHR_DiseaseDuration 0.542798
## ms_dx_condensedProgressive MS 0.172629
## race_ethnicity_cleanBlack Or African American 0.034106 *
## race_ethnicity_cleanHispanic or Latino 0.517087
## race_ethnicity_cleanWhite Not Hispanic 0.505968
## clean_sexMale 0.977307
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2578 on 49 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.3129, Adjusted R-squared: 0.2007
## F-statistic: 2.789 on 8 and 49 DF, p-value: 0.01258
# no pressure mat value
# unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, foot1_stance_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 122 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.36932 -0.24090 -0.02433 0.16763 0.77262
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.9217 0.2814 3.276 0.00281 **
## foot1_stance_time_mean_pose_zv 1.0732 0.3726 2.880 0.00754 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.284 on 28 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.2285, Adjusted R-squared: 0.201
## F-statistic: 8.295 on 1 and 28 DF, p-value: 0.00754
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.3220 -0.1738 -0.0485 0.1697 0.6210
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.165880 0.372639 3.129
## foot1_stance_time_mean_pose_zv 0.990039 0.402816 2.458
## demoEHR_Age -0.001072 0.005007 -0.214
## demoEHR_DiseaseDuration 0.005048 0.007126 0.708
## ms_dx_condensedProgressive MS 0.179869 0.166100 1.083
## race_ethnicity_cleanBlack Or African American 0.326953 0.261415 1.251
## race_ethnicity_cleanHispanic or Latino -0.152081 0.210890 -0.721
## race_ethnicity_cleanWhite Not Hispanic -0.268644 0.190168 -1.413
## clean_sexMale -0.085978 0.131760 -0.653
## Pr(>|t|)
## (Intercept) 0.00508 **
## foot1_stance_time_mean_pose_zv 0.02276 *
## demoEHR_Age 0.83254
## demoEHR_DiseaseDuration 0.48646
## ms_dx_condensedProgressive MS 0.29113
## race_ethnicity_cleanBlack Or African American 0.22480
## race_ethnicity_cleanHispanic or Latino 0.47878
## race_ethnicity_cleanWhite Not Hispanic 0.17240
## clean_sexMale 0.52114
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2712 on 21 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.4727, Adjusted R-squared: 0.2718
## F-statistic: 2.353 on 8 and 21 DF, p-value: 0.05555
Univariate - PWS, Swing Time –> T25fw
# PWS
sum(is.finite(zeno_pws_df$foot1_swing_time_mean_pose_zv))## [1] 58ggplot(data = zeno_pws_df, aes(x = foot1_swing_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_swing_per_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
#video model
metric_regression(zeno_pws_df, t25fw_log, foot1_swing_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 159 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_swing_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.45529 -0.16233 -0.04102 0.12953 1.09583
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.3620 0.1905 7.148 1.97e-09 ***
## foot1_swing_time_mean_pose_zv 0.6748 0.5059 1.334 0.188
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2864 on 56 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.03079, Adjusted R-squared: 0.01349
## F-statistic: 1.779 on 1 and 56 DF, p-value: 0.1876
## [1] "t25fw_log ~ foot1_swing_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.46218 -0.12759 -0.02431 0.12562 0.87576
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.306485 0.273642 4.774
## foot1_swing_time_mean_pose_zv 0.419610 0.503006 0.834
## demoEHR_Age 0.002587 0.004215 0.614
## demoEHR_DiseaseDuration 0.002093 0.005635 0.372
## ms_dx_condensedProgressive MS 0.255542 0.133859 1.909
## race_ethnicity_cleanBlack Or African American 0.410745 0.202207 2.031
## race_ethnicity_cleanHispanic or Latino -0.086843 0.143656 -0.605
## race_ethnicity_cleanWhite Not Hispanic -0.060207 0.132270 -0.455
## clean_sexMale 0.036508 0.086870 0.420
## Pr(>|t|)
## (Intercept) 1.67e-05 ***
## foot1_swing_time_mean_pose_zv 0.4082
## demoEHR_Age 0.5422
## demoEHR_DiseaseDuration 0.7119
## ms_dx_condensedProgressive MS 0.0621 .
## race_ethnicity_cleanBlack Or African American 0.0477 *
## race_ethnicity_cleanHispanic or Latino 0.5483
## race_ethnicity_cleanWhite Not Hispanic 0.6510
## clean_sexMale 0.6761
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2693 on 49 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.2501, Adjusted R-squared: 0.1277
## F-statistic: 2.043 on 8 and 49 DF, p-value: 0.06047
# no pressure mat value
# unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, foot1_swing_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 122 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_swing_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.42629 -0.18620 -0.07886 0.11899 0.96976
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.5843 0.2696 5.877 2.55e-06 ***
## foot1_swing_time_mean_pose_zv 0.3540 0.6953 0.509 0.615
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3219 on 28 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.009169, Adjusted R-squared: -0.02622
## F-statistic: 0.2591 on 1 and 28 DF, p-value: 0.6147
## [1] "t25fw_log ~ foot1_swing_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.34769 -0.20452 -0.03124 0.11781 0.69450
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.7590522 0.4453506 3.950
## foot1_swing_time_mean_pose_zv -0.0110863 0.7343771 -0.015
## demoEHR_Age 0.0005365 0.0056348 0.095
## demoEHR_DiseaseDuration 0.0037865 0.0081471 0.465
## ms_dx_condensedProgressive MS 0.3281831 0.1760724 1.864
## race_ethnicity_cleanBlack Or African American 0.3877825 0.2981793 1.301
## race_ethnicity_cleanHispanic or Latino -0.1017299 0.2389752 -0.426
## race_ethnicity_cleanWhite Not Hispanic -0.2277727 0.2173307 -1.048
## clean_sexMale -0.0160778 0.1461782 -0.110
## Pr(>|t|)
## (Intercept) 0.000732 ***
## foot1_swing_time_mean_pose_zv 0.988098
## demoEHR_Age 0.925054
## demoEHR_DiseaseDuration 0.646878
## ms_dx_condensedProgressive MS 0.076377 .
## race_ethnicity_cleanBlack Or African American 0.207521
## race_ethnicity_cleanHispanic or Latino 0.674663
## race_ethnicity_cleanWhite Not Hispanic 0.306530
## clean_sexMale 0.913464
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3077 on 21 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.321, Adjusted R-squared: 0.06234
## F-statistic: 1.241 on 8 and 21 DF, p-value: 0.3246
Univariate - PWS, Single and Double Support –> T25fw
# PWS
sum(is.finite(zeno_pws_df$foot1_double_support_per_mean_pose_zv))## [1] 58ggplot(data = zeno_pws_df, aes(x = foot1_double_support_per_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_ini_double_support_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_single_support_per_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_term_double_support_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = foot1_tot_double_support_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 159 rows containing missing values (`geom_point()`).
# try terminal double support time
metric_regression(zeno_pws_df, t25fw_log, foot1_term_double_support_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 159 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.46652 -0.13854 -0.04039 0.16264 0.87022
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.40581 0.07162 19.628
## foot1_term_double_support_time_mean_pose_zv 0.96059 0.29240 3.285
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## foot1_term_double_support_time_mean_pose_zv 0.00176 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2664 on 56 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.1616, Adjusted R-squared: 0.1466
## F-statistic: 10.79 on 1 and 56 DF, p-value: 0.001762
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47253 -0.13502 -0.03502 0.10785 0.78821
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.298238 0.204473 6.349
## foot1_term_double_support_time_mean_pose_zv 0.742819 0.322892 2.301
## demoEHR_Age 0.002214 0.004030 0.549
## demoEHR_DiseaseDuration 0.005164 0.005536 0.933
## ms_dx_condensedProgressive MS 0.131781 0.137203 0.960
## race_ethnicity_cleanBlack Or African American 0.347911 0.193138 1.801
## race_ethnicity_cleanHispanic or Latino -0.047579 0.137428 -0.346
## race_ethnicity_cleanWhite Not Hispanic -0.066865 0.126586 -0.528
## clean_sexMale 0.042009 0.082956 0.506
## Pr(>|t|)
## (Intercept) 6.76e-08 ***
## foot1_term_double_support_time_mean_pose_zv 0.0257 *
## demoEHR_Age 0.5852
## demoEHR_DiseaseDuration 0.3555
## ms_dx_condensedProgressive MS 0.3415
## race_ethnicity_cleanBlack Or African American 0.0778 .
## race_ethnicity_cleanHispanic or Latino 0.7307
## race_ethnicity_cleanWhite Not Hispanic 0.5997
## clean_sexMale 0.6148
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2576 on 49 degrees of freedom
## (159 observations deleted due to missingness)
## Multiple R-squared: 0.3136, Adjusted R-squared: 0.2015
## F-statistic: 2.798 on 8 and 49 DF, p-value: 0.01234
# unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, foot1_term_double_support_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 122 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47957 -0.21192 -0.01916 0.19843 0.75915
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.48305 0.08981 16.513
## foot1_term_double_support_time_mean_pose_zv 1.04515 0.32959 3.171
## Pr(>|t|)
## (Intercept) 5.77e-16 ***
## foot1_term_double_support_time_mean_pose_zv 0.00366 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2774 on 28 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.2642, Adjusted R-squared: 0.238
## F-statistic: 10.06 on 1 and 28 DF, p-value: 0.003664
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.29714 -0.19772 -0.03331 0.10630 0.63890
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.6517784 0.3007771 5.492
## foot1_term_double_support_time_mean_pose_zv 0.8110046 0.4014502 2.020
## demoEHR_Age -0.0007846 0.0051959 -0.151
## demoEHR_DiseaseDuration 0.0059087 0.0074536 0.793
## ms_dx_condensedProgressive MS 0.1760955 0.1774567 0.992
## race_ethnicity_cleanBlack Or African American 0.2183771 0.2828617 0.772
## race_ethnicity_cleanHispanic or Latino -0.1146077 0.2180385 -0.526
## race_ethnicity_cleanWhite Not Hispanic -0.2338653 0.1967099 -1.189
## clean_sexMale -0.0104773 0.1336019 -0.078
## Pr(>|t|)
## (Intercept) 1.9e-05 ***
## foot1_term_double_support_time_mean_pose_zv 0.0563 .
## demoEHR_Age 0.8814
## demoEHR_DiseaseDuration 0.4368
## ms_dx_condensedProgressive MS 0.3323
## race_ethnicity_cleanBlack Or African American 0.4487
## race_ethnicity_cleanHispanic or Latino 0.6047
## race_ethnicity_cleanWhite Not Hispanic 0.2478
## clean_sexMale 0.9382
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2816 on 21 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.4315, Adjusted R-squared: 0.2149
## F-statistic: 1.992 on 8 and 21 DF, p-value: 0.09835
# mat double support
metric_regression(zeno_pws_df, t25fw_log, PWS_totaldsupportmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ PWS_totaldsupportmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.12428 -0.20113 -0.05144 0.12347 1.24210
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.887291 0.061276 14.48 <2e-16 ***
## PWS_totaldsupportmean 0.022743 0.001745 13.04 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3315 on 215 degrees of freedom
## Multiple R-squared: 0.4414, Adjusted R-squared: 0.4388
## F-statistic: 169.9 on 1 and 215 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ PWS_totaldsupportmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.78132 -0.15720 -0.04603 0.12134 1.14323
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.855723 0.121922 7.019
## PWS_totaldsupportmean 0.019008 0.001789 10.625
## demoEHR_Age 0.006192 0.002297 2.695
## demoEHR_DiseaseDuration -0.002157 0.003026 -0.713
## ms_dx_condensedProgressive MS 0.224710 0.066436 3.382
## ms_dx_condensedMS, Subtype Not Specified -0.003515 0.224517 -0.016
## race_ethnicity_cleanBlack Or African American 0.049273 0.115178 0.428
## race_ethnicity_cleanHispanic or Latino -0.119580 0.103274 -1.158
## race_ethnicity_cleanWhite Not Hispanic -0.205655 0.084148 -2.444
## race_ethnicity_cleanOther/Unknown/Declined -0.131118 0.106288 -1.234
## clean_sexMale -0.053597 0.051050 -1.050
## clean_sexNon-Binary 0.036730 0.222890 0.165
## Pr(>|t|)
## (Intercept) 3.22e-11 ***
## PWS_totaldsupportmean < 2e-16 ***
## demoEHR_Age 0.007614 **
## demoEHR_DiseaseDuration 0.476884
## ms_dx_condensedProgressive MS 0.000861 ***
## ms_dx_condensedMS, Subtype Not Specified 0.987526
## race_ethnicity_cleanBlack Or African American 0.669247
## race_ethnicity_cleanHispanic or Latino 0.248258
## race_ethnicity_cleanWhite Not Hispanic 0.015372 *
## race_ethnicity_cleanOther/Unknown/Declined 0.218759
## clean_sexMale 0.295005
## clean_sexNon-Binary 0.869271
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3097 on 205 degrees of freedom
## Multiple R-squared: 0.5352, Adjusted R-squared: 0.5102
## F-statistic: 21.46 on 11 and 205 DF, p-value: < 2.2e-16
zeno_pws_df$log_PWS_totaldsupportmean <- log(zeno_pws_df$PWS_totaldsupportmean)
metric_regression(zeno_pws_df, t25fw_log, log_PWS_totaldsupportmean)## [1] "Data Frame: zeno_pws_df"
## [1] "t25fw_log ~ log_PWS_totaldsupportmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.38174 -0.17182 -0.02423 0.15015 0.88946
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.64659 0.27590 -13.22 <2e-16 ***
## log_PWS_totaldsupportmean 1.52906 0.07977 19.17 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2695 on 215 degrees of freedom
## Multiple R-squared: 0.6308, Adjusted R-squared: 0.6291
## F-statistic: 367.4 on 1 and 215 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_PWS_totaldsupportmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.3059 -0.1497 -0.0212 0.1166 0.8261
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) -3.275760 0.302932 -10.814
## log_PWS_totaldsupportmean 1.413987 0.088181 16.035
## demoEHR_Age 0.005156 0.001908 2.702
## demoEHR_DiseaseDuration -0.002041 0.002508 -0.814
## ms_dx_condensedProgressive MS 0.101121 0.056559 1.788
## ms_dx_condensedMS, Subtype Not Specified 0.038323 0.186249 0.206
## race_ethnicity_cleanBlack Or African American -0.134514 0.096995 -1.387
## race_ethnicity_cleanHispanic or Latino -0.174660 0.085826 -2.035
## race_ethnicity_cleanWhite Not Hispanic -0.241929 0.069888 -3.462
## race_ethnicity_cleanOther/Unknown/Declined -0.181906 0.088285 -2.060
## clean_sexMale -0.095291 0.042432 -2.246
## clean_sexNon-Binary 0.022332 0.184851 0.121
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## log_PWS_totaldsupportmean < 2e-16 ***
## demoEHR_Age 0.007467 **
## demoEHR_DiseaseDuration 0.416689
## ms_dx_condensedProgressive MS 0.075270 .
## ms_dx_condensedMS, Subtype Not Specified 0.837180
## race_ethnicity_cleanBlack Or African American 0.167005
## race_ethnicity_cleanHispanic or Latino 0.043132 *
## race_ethnicity_cleanWhite Not Hispanic 0.000653 ***
## race_ethnicity_cleanOther/Unknown/Declined 0.040619 *
## clean_sexMale 0.025789 *
## clean_sexNon-Binary 0.903958
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2569 on 205 degrees of freedom
## Multiple R-squared: 0.6803, Adjusted R-squared: 0.6631
## F-statistic: 39.65 on 11 and 205 DF, p-value: < 2.2e-16
Multivariate - PWS, Gait metrics only –> T25fw
Metrics only - not including double support/stance measures, too many missing. May include after improving code
# PWS
# confounding +
pws_t25fw_multivar_model <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.79102 -0.20897 -0.04041 0.13623 1.24196
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.698883 0.520653 1.342 0.18149
## log_delta_pix_h_rel_median_pose_zv -0.179348 0.058299 -3.076 0.00249 **
## stride_time_median_sec_pose_zv 0.480250 0.233544 2.056 0.04146 *
## mean_cadence_step_per_min_pose_zv -0.002185 0.002759 -0.792 0.42969
## stride_width_median_cm_pose_zv 0.027336 0.007860 3.478 0.00066 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3095 on 152 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.2844, Adjusted R-squared: 0.2655
## F-statistic: 15.1 on 4 and 152 DF, p-value: 2.048e-10hist(resid(pws_t25fw_multivar_model))
# PWS interaction *
pws_t25fw_multivar_model_2 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv *
stride_time_median_sec_pose_zv *
mean_cadence_step_per_min_pose_zv *
stride_width_median_cm_pose_zv,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model_2)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv *
## stride_time_median_sec_pose_zv * mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.68168 -0.17935 -0.02735 0.13538 1.19404
##
## Coefficients:
## Estimate
## (Intercept) -19.950547
## log_delta_pix_h_rel_median_pose_zv -1.898243
## stride_time_median_sec_pose_zv 19.453924
## mean_cadence_step_per_min_pose_zv 0.183459
## stride_width_median_cm_pose_zv 2.099951
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 3.140075
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.019423
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.168232
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.486981
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -1.852010
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.017724
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.029751
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -0.519144
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.004034
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.015654
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.004216
## Std. Error
## (Intercept) 15.280293
## log_delta_pix_h_rel_median_pose_zv 10.748354
## stride_time_median_sec_pose_zv 11.755009
## mean_cadence_step_per_min_pose_zv 0.128764
## stride_width_median_cm_pose_zv 1.220568
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 7.253678
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.098987
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.100479
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.799679
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.995378
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.010540
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.070700
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.546177
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.007378
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.008756
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.005222
## t value
## (Intercept) -1.306
## log_delta_pix_h_rel_median_pose_zv -0.177
## stride_time_median_sec_pose_zv 1.655
## mean_cadence_step_per_min_pose_zv 1.425
## stride_width_median_cm_pose_zv 1.720
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 0.433
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.196
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -1.674
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.609
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -1.861
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -1.682
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.421
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -0.951
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.547
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 1.788
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.807
## Pr(>|t|)
## (Intercept) 0.1938
## log_delta_pix_h_rel_median_pose_zv 0.8601
## stride_time_median_sec_pose_zv 0.1002
## mean_cadence_step_per_min_pose_zv 0.1564
## stride_width_median_cm_pose_zv 0.0875
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 0.6658
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.8447
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.0963
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.5435
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.0649
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.0949
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.6745
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.3435
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.5854
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.0760
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.4208
##
## (Intercept)
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv
## mean_cadence_step_per_min_pose_zv
## stride_width_median_cm_pose_zv .
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv .
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv .
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv .
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv .
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2745 on 141 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4779, Adjusted R-squared: 0.4223
## F-statistic: 8.603 on 15 and 141 DF, p-value: 8.904e-14hist(resid(pws_t25fw_multivar_model_2))
Multivariate - PWS, Gait metrics + disease + demographics –> T25fw
# PWS
# Metrics + disease and demographic info
# add MS subtype
pws_t25fw_multivar_model_3 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model_3)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.76721 -0.19501 -0.04181 0.14730 1.19160
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.890449 0.525961 1.693 0.09253
## log_delta_pix_h_rel_median_pose_zv -0.155942 0.059180 -2.635 0.00930
## stride_time_median_sec_pose_zv 0.408772 0.234663 1.742 0.08357
## mean_cadence_step_per_min_pose_zv -0.002526 0.002749 -0.919 0.35965
## stride_width_median_cm_pose_zv 0.022193 0.008228 2.697 0.00779
## ms_dx_condensedProgressive MS 0.166632 0.082591 2.018 0.04542
## ms_dx_condensedMS, Subtype Not Specified 0.031242 0.220317 0.142 0.88742
##
## (Intercept) .
## log_delta_pix_h_rel_median_pose_zv **
## stride_time_median_sec_pose_zv .
## mean_cadence_step_per_min_pose_zv
## stride_width_median_cm_pose_zv **
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3074 on 150 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.3033, Adjusted R-squared: 0.2754
## F-statistic: 10.88 on 6 and 150 DF, p-value: 4.851e-10hist(resid(pws_t25fw_multivar_model_3))
# add age
pws_t25fw_multivar_model_4 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model_4)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age,
## data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.76560 -0.18098 -0.03939 0.11653 1.19800
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.516580 0.528061 0.978 0.32953
## log_delta_pix_h_rel_median_pose_zv -0.171593 0.057940 -2.962 0.00356
## stride_time_median_sec_pose_zv 0.460465 0.229453 2.007 0.04658
## mean_cadence_step_per_min_pose_zv -0.002597 0.002680 -0.969 0.33407
## stride_width_median_cm_pose_zv 0.022179 0.008022 2.765 0.00641
## ms_dx_condensedProgressive MS 0.083888 0.085219 0.984 0.32652
## ms_dx_condensedMS, Subtype Not Specified -0.086723 0.218454 -0.397 0.69195
## demoEHR_Age 0.006334 0.002135 2.966 0.00351
##
## (Intercept)
## log_delta_pix_h_rel_median_pose_zv **
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv
## stride_width_median_cm_pose_zv **
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2997 on 149 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.3421, Adjusted R-squared: 0.3112
## F-statistic: 11.07 on 7 and 149 DF, p-value: 3.155e-11hist(resid(pws_t25fw_multivar_model_4))
# add disease duration
pws_t25fw_multivar_model_5 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model_5)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.75055 -0.20106 -0.03773 0.12279 1.22751
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.537966 0.529577 1.016 0.31136
## log_delta_pix_h_rel_median_pose_zv -0.167871 0.058231 -2.883 0.00453
## stride_time_median_sec_pose_zv 0.476349 0.230742 2.064 0.04072
## mean_cadence_step_per_min_pose_zv -0.002703 0.002688 -1.006 0.31614
## stride_width_median_cm_pose_zv 0.021763 0.008052 2.703 0.00768
## ms_dx_condensedProgressive MS 0.078772 0.085609 0.920 0.35900
## ms_dx_condensedMS, Subtype Not Specified -0.090385 0.218822 -0.413 0.68016
## demoEHR_Age 0.005436 0.002446 2.222 0.02780
## demoEHR_DiseaseDuration 0.002653 0.003509 0.756 0.45077
##
## (Intercept)
## log_delta_pix_h_rel_median_pose_zv **
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv
## stride_width_median_cm_pose_zv **
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age *
## demoEHR_DiseaseDuration
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3001 on 148 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.3447, Adjusted R-squared: 0.3092
## F-statistic: 9.729 on 8 and 148 DF, p-value: 8.482e-11hist(resid(pws_t25fw_multivar_model_5))
# add race and ethnicity
pws_t25fw_multivar_model_6 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = zeno_pws_df)
summary(pws_t25fw_multivar_model_6)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.69971 -0.16441 -0.02223 0.11795 1.18667
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.841785 0.521249 1.615
## log_delta_pix_h_rel_median_pose_zv -0.164947 0.055873 -2.952
## stride_time_median_sec_pose_zv 0.360265 0.218411 1.649
## mean_cadence_step_per_min_pose_zv -0.003413 0.002573 -1.326
## stride_width_median_cm_pose_zv 0.014601 0.007732 1.888
## ms_dx_condensedProgressive MS 0.129769 0.080907 1.604
## ms_dx_condensedMS, Subtype Not Specified -0.040427 0.205264 -0.197
## demoEHR_Age 0.006561 0.002446 2.682
## demoEHR_DiseaseDuration 0.002179 0.003292 0.662
## race_ethnicity_cleanBlack Or African American 0.317858 0.124205 2.559
## race_ethnicity_cleanHispanic or Latino -0.052037 0.117537 -0.443
## race_ethnicity_cleanWhite Not Hispanic -0.115362 0.098188 -1.175
## race_ethnicity_cleanOther/Unknown/Declined -0.018505 0.132571 -0.140
## Pr(>|t|)
## (Intercept) 0.10851
## log_delta_pix_h_rel_median_pose_zv 0.00369 **
## stride_time_median_sec_pose_zv 0.10123
## mean_cadence_step_per_min_pose_zv 0.18691
## stride_width_median_cm_pose_zv 0.06099 .
## ms_dx_condensedProgressive MS 0.11092
## ms_dx_condensedMS, Subtype Not Specified 0.84414
## demoEHR_Age 0.00818 **
## demoEHR_DiseaseDuration 0.50900
## race_ethnicity_cleanBlack Or African American 0.01153 *
## race_ethnicity_cleanHispanic or Latino 0.65863
## race_ethnicity_cleanWhite Not Hispanic 0.24197
## race_ethnicity_cleanOther/Unknown/Declined 0.88919
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.281 on 144 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4409, Adjusted R-squared: 0.3943
## F-statistic: 9.463 on 12 and 144 DF, p-value: 2.374e-13hist(resid(pws_t25fw_multivar_model_6))
Fast Walking Speed –> T25fw
Demographics and disease info
# Fast Walking speed participants
fw_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
race_ethnicity_clean +
clean_sex,
data = zeno_fw_df)
summary(fw_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + race_ethnicity_clean + clean_sex, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.04904 -0.21155 -0.05219 0.13159 1.78029
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.337602 0.139803 9.568
## demoEHR_Age 0.007637 0.002837 2.692
## demoEHR_DiseaseDuration -0.004296 0.003735 -1.150
## ms_dx_condensedProgressive MS 0.464139 0.077080 6.021
## ms_dx_condensedMS, Subtype Not Specified -0.035212 0.278365 -0.126
## race_ethnicity_cleanBlack Or African American 0.179691 0.141982 1.266
## race_ethnicity_cleanHispanic or Latino -0.061529 0.127929 -0.481
## race_ethnicity_cleanWhite Not Hispanic -0.160184 0.104237 -1.537
## race_ethnicity_cleanOther/Unknown/Declined -0.088865 0.131740 -0.675
## clean_sexMale -0.051636 0.063302 -0.816
## clean_sexNon-Binary 0.007643 0.276484 0.028
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## demoEHR_Age 0.00768 **
## demoEHR_DiseaseDuration 0.25131
## ms_dx_condensedProgressive MS 7.77e-09 ***
## ms_dx_condensedMS, Subtype Not Specified 0.89946
## race_ethnicity_cleanBlack Or African American 0.20708
## race_ethnicity_cleanHispanic or Latino 0.63105
## race_ethnicity_cleanWhite Not Hispanic 0.12589
## race_ethnicity_cleanOther/Unknown/Declined 0.50071
## clean_sexMale 0.41561
## clean_sexNon-Binary 0.97797
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3842 on 207 degrees of freedom
## Multiple R-squared: 0.2786, Adjusted R-squared: 0.2437
## F-statistic: 7.993 on 10 and 207 DF, p-value: 7.383e-11hist(resid(fw_dem_model))
# unique IDs
fw_dem_model_2 <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
ms_dx_condensed +
clean_sex,
data = zeno_fw_uniqueid_df)
summary(fw_dem_model_2)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + ms_dx_condensed + clean_sex, data = zeno_fw_uniqueid_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.74546 -0.21175 -0.05542 0.13677 1.56730
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.362464 0.134571 10.124 < 2e-16
## demoEHR_Age 0.004441 0.002933 1.514 0.132
## demoEHR_DiseaseDuration -0.002086 0.004571 -0.456 0.649
## ms_dx_condensedProgressive MS 0.504910 0.086688 5.824 3.55e-08
## ms_dx_condensedMS, Subtype Not Specified -0.055911 0.275970 -0.203 0.840
## clean_sexMale -0.028488 0.071505 -0.398 0.691
## clean_sexNon-Binary -0.062676 0.382089 -0.164 0.870
##
## (Intercept) ***
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS ***
## ms_dx_condensedMS, Subtype Not Specified
## clean_sexMale
## clean_sexNon-Binary
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3783 on 145 degrees of freedom
## Multiple R-squared: 0.2571, Adjusted R-squared: 0.2264
## F-statistic: 8.364 on 6 and 145 DF, p-value: 8.591e-08hist(resid(fw_dem_model_2))
Univariate - FW, delta pixel proxy –> T25fw
sum(is.finite(zeno_fw_df$delta_pix_h_rel_median_pose_zv))## [1] 215zeno_fw_df$log_delta_pix_h_rel_median_pose_zv <-log(zeno_fw_df$delta_pix_h_rel_median_pose_zv)
zeno_fw_df$sqrt_delta_pix_h_rel_median_pose_zv <- sqrt(zeno_fw_df$delta_pix_h_rel_median_pose_zv)
# drop
zeno_fw_df <- zeno_fw_df %>% filter(is.finite(log_delta_pix_h_rel_median_pose_zv))
nrow(zeno_fw_df)## [1] 214# log T25FW
ggplot(data = zeno_fw_df, aes(x = delta_pix_h_rel_median_pose_zv, y = t25fw_log)) +
geom_point()
ggplot(data = zeno_fw_df, aes(x = log_delta_pix_h_rel_median_pose_zv, y = t25fw_log)) +
geom_point()
ggplot(data = zeno_fw_df, aes(x = sqrt_delta_pix_h_rel_median_pose_zv, y = t25fw_log)) +
geom_point()
# log velproxy and log t25Fw
metric_regression(zeno_fw_df, t25fw_log, log_delta_pix_h_rel_median_pose_zv)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.26427 -0.18434 -0.03673 0.12861 1.68174
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.06395 0.04933 21.57 <2e-16 ***
## log_delta_pix_h_rel_median_pose_zv -0.53911 0.04142 -13.02 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3328 on 212 degrees of freedom
## Multiple R-squared: 0.4442, Adjusted R-squared: 0.4415
## F-statistic: 169.4 on 1 and 212 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.98030 -0.16891 -0.00078 0.13002 1.42428
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.982743 0.121667 8.077
## log_delta_pix_h_rel_median_pose_zv -0.442881 0.042902 -10.323
## demoEHR_Age 0.005034 0.002359 2.134
## demoEHR_DiseaseDuration -0.003905 0.003096 -1.261
## ms_dx_condensedProgressive MS 0.260872 0.066180 3.942
## ms_dx_condensedMS, Subtype Not Specified 0.259071 0.322729 0.803
## race_ethnicity_cleanBlack Or African American 0.117963 0.117509 1.004
## race_ethnicity_cleanHispanic or Latino -0.098009 0.107737 -0.910
## race_ethnicity_cleanWhite Not Hispanic -0.093375 0.087275 -1.070
## race_ethnicity_cleanOther/Unknown/Declined -0.059841 0.109121 -0.548
## clean_sexMale -0.027375 0.053168 -0.515
## clean_sexNon-Binary 0.050453 0.226106 0.223
## Pr(>|t|)
## (Intercept) 5.91e-14 ***
## log_delta_pix_h_rel_median_pose_zv < 2e-16 ***
## demoEHR_Age 0.034025 *
## demoEHR_DiseaseDuration 0.208680
## ms_dx_condensedProgressive MS 0.000111 ***
## ms_dx_condensedMS, Subtype Not Specified 0.423062
## race_ethnicity_cleanBlack Or African American 0.316646
## race_ethnicity_cleanHispanic or Latino 0.364063
## race_ethnicity_cleanWhite Not Hispanic 0.285943
## race_ethnicity_cleanOther/Unknown/Declined 0.584028
## clean_sexMale 0.607208
## clean_sexNon-Binary 0.823652
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3141 on 202 degrees of freedom
## Multiple R-squared: 0.5284, Adjusted R-squared: 0.5028
## F-statistic: 20.58 on 11 and 202 DF, p-value: < 2.2e-16
# true FW velocity from mat and t25fw
metric_regression(zeno_fw_df, t25fw_log, FW_velocitycmsecmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ FW_velocitycmsecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.61784 -0.10610 -0.02210 0.09633 1.30904
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.874161 0.057933 49.61 <2e-16 ***
## FW_velocitycmsecmean -0.008315 0.000372 -22.36 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2436 on 212 degrees of freedom
## Multiple R-squared: 0.7022, Adjusted R-squared: 0.7008
## F-statistic: 499.8 on 1 and 212 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ FW_velocitycmsecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.61197 -0.11747 -0.01485 0.10063 1.23481
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.8621272 0.1267148 22.587
## FW_velocitycmsecmean -0.0080401 0.0004630 -17.366
## demoEHR_Age -0.0009546 0.0019017 -0.502
## demoEHR_DiseaseDuration -0.0015267 0.0024297 -0.628
## ms_dx_condensedProgressive MS 0.1146547 0.0533962 2.147
## ms_dx_condensedMS, Subtype Not Specified 0.0756038 0.2527457 0.299
## race_ethnicity_cleanBlack Or African American 0.0102379 0.0923861 0.111
## race_ethnicity_cleanHispanic or Latino -0.0027676 0.0843402 -0.033
## race_ethnicity_cleanWhite Not Hispanic 0.0075481 0.0688067 0.110
## race_ethnicity_cleanOther/Unknown/Declined 0.0518258 0.0857746 0.604
## clean_sexMale 0.0094030 0.0417386 0.225
## clean_sexNon-Binary 0.1250206 0.1770966 0.706
## Pr(>|t|)
## (Intercept) <2e-16 ***
## FW_velocitycmsecmean <2e-16 ***
## demoEHR_Age 0.616
## demoEHR_DiseaseDuration 0.530
## ms_dx_condensedProgressive MS 0.033 *
## ms_dx_condensedMS, Subtype Not Specified 0.765
## race_ethnicity_cleanBlack Or African American 0.912
## race_ethnicity_cleanHispanic or Latino 0.974
## race_ethnicity_cleanWhite Not Hispanic 0.913
## race_ethnicity_cleanOther/Unknown/Declined 0.546
## clean_sexMale 0.822
## clean_sexNon-Binary 0.481
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2458 on 202 degrees of freedom
## Multiple R-squared: 0.711, Adjusted R-squared: 0.6953
## F-statistic: 45.19 on 11 and 202 DF, p-value: < 2.2e-16
# unique IDs
zeno_fw_uniqueid_df$log_delta_pix_h_rel_median_pose_zv <- log(zeno_fw_uniqueid_df$delta_pix_h_rel_median_pose_zv)
zeno_fw_uniqueid_df[] <- lapply(zeno_fw_uniqueid_df, function(x) {
if (is.numeric(x)) replace(x, is.infinite(x), NA) else x
})
metric_regression(zeno_fw_uniqueid_df, t25fw_log, log_delta_pix_h_rel_median_pose_zv)## [1] "Data Frame: zeno_fw_uniqueid_df"## Warning: Removed 2 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.35233 -0.17374 -0.02114 0.13592 1.65013
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.06175 0.05603 18.95 <2e-16 ***
## log_delta_pix_h_rel_median_pose_zv -0.58389 0.04958 -11.78 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3115 on 148 degrees of freedom
## (2 observations deleted due to missingness)
## Multiple R-squared: 0.4838, Adjusted R-squared: 0.4803
## F-statistic: 138.7 on 1 and 148 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.11570 -0.12047 -0.01092 0.10797 1.37698
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.044096 0.137238 7.608
## log_delta_pix_h_rel_median_pose_zv -0.498234 0.055946 -8.906
## demoEHR_Age 0.003640 0.002509 1.451
## demoEHR_DiseaseDuration -0.003502 0.003716 -0.942
## ms_dx_condensedProgressive MS 0.236181 0.075157 3.143
## ms_dx_condensedMS, Subtype Not Specified 0.231689 0.313889 0.738
## race_ethnicity_cleanBlack Or African American -0.017933 0.130692 -0.137
## race_ethnicity_cleanHispanic or Latino -0.134506 0.122976 -1.094
## race_ethnicity_cleanWhite Not Hispanic -0.110564 0.099751 -1.108
## race_ethnicity_cleanOther/Unknown/Declined 0.031390 0.128366 0.245
## clean_sexMale -0.008757 0.059583 -0.147
## clean_sexNon-Binary -0.092261 0.306123 -0.301
## Pr(>|t|)
## (Intercept) 3.92e-12 ***
## log_delta_pix_h_rel_median_pose_zv 2.74e-15 ***
## demoEHR_Age 0.14915
## demoEHR_DiseaseDuration 0.34763
## ms_dx_condensedProgressive MS 0.00205 **
## ms_dx_condensedMS, Subtype Not Specified 0.46169
## race_ethnicity_cleanBlack Or African American 0.89106
## race_ethnicity_cleanHispanic or Latino 0.27597
## race_ethnicity_cleanWhite Not Hispanic 0.26962
## race_ethnicity_cleanOther/Unknown/Declined 0.80718
## clean_sexMale 0.88337
## clean_sexNon-Binary 0.76357
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.302 on 138 degrees of freedom
## (2 observations deleted due to missingness)
## Multiple R-squared: 0.5476, Adjusted R-squared: 0.5115
## F-statistic: 15.18 on 11 and 138 DF, p-value: < 2.2e-16
Univariate - FW, stride time –> T25fw
sum(is.finite(zeno_fw_df$stride_time_mean_sec_pose_zv))## [1] 163metric_regression(zeno_fw_df, t25fw_log, stride_time_median_sec_pose_zv)## [1] "Data Frame: zeno_fw_df"## Warning: Removed 51 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.73060 -0.18730 -0.04261 0.12506 1.92101
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.2280 0.1508 1.512 0.133
## stride_time_median_sec_pose_zv 1.4712 0.1574 9.348 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3402 on 161 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.3518, Adjusted R-squared: 0.3478
## F-statistic: 87.38 on 1 and 161 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ stride_time_median_sec_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.64025 -0.15865 -0.03891 0.11763 1.59607
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.213514 0.206204 1.035
## stride_time_median_sec_pose_zv 1.218923 0.161245 7.559
## demoEHR_Age 0.006493 0.002614 2.484
## demoEHR_DiseaseDuration 0.002095 0.003583 0.585
## ms_dx_condensedProgressive MS 0.197327 0.082560 2.390
## ms_dx_condensedMS, Subtype Not Specified 0.227910 0.324963 0.701
## race_ethnicity_cleanBlack Or African American 0.213650 0.128212 1.666
## race_ethnicity_cleanHispanic or Latino -0.033581 0.119039 -0.282
## race_ethnicity_cleanWhite Not Hispanic -0.158871 0.097574 -1.628
## race_ethnicity_cleanOther/Unknown/Declined -0.080149 0.127595 -0.628
## clean_sexMale -0.089532 0.058921 -1.520
## clean_sexNon-Binary 0.045792 0.228002 0.201
## Pr(>|t|)
## (Intercept) 0.3021
## stride_time_median_sec_pose_zv 3.64e-12 ***
## demoEHR_Age 0.0141 *
## demoEHR_DiseaseDuration 0.5597
## ms_dx_condensedProgressive MS 0.0181 *
## ms_dx_condensedMS, Subtype Not Specified 0.4842
## race_ethnicity_cleanBlack Or African American 0.0977 .
## race_ethnicity_cleanHispanic or Latino 0.7783
## race_ethnicity_cleanWhite Not Hispanic 0.1056
## race_ethnicity_cleanOther/Unknown/Declined 0.5309
## clean_sexMale 0.1307
## clean_sexNon-Binary 0.8411
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3145 on 151 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.4807, Adjusted R-squared: 0.4429
## F-statistic: 12.71 on 11 and 151 DF, p-value: < 2.2e-16
# true fast walk - stride time from mat
metric_regression(zeno_fw_df, t25fw_log, FW_stridetimesecmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ FW_stridetimesecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.57048 -0.16524 -0.01128 0.15425 0.82409
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.20303 0.06827 2.974 0.00328 **
## FW_stridetimesecmean 1.42136 0.06568 21.642 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2492 on 212 degrees of freedom
## Multiple R-squared: 0.6884, Adjusted R-squared: 0.6869
## F-statistic: 468.4 on 1 and 212 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ FW_stridetimesecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.49823 -0.13689 -0.00533 0.12817 0.82342
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.171658 0.106602 1.610
## FW_stridetimesecmean 1.337358 0.070935 18.853
## demoEHR_Age 0.004395 0.001753 2.507
## demoEHR_DiseaseDuration 0.003300 0.002341 1.409
## ms_dx_condensedProgressive MS 0.067288 0.051507 1.306
## ms_dx_condensedMS, Subtype Not Specified 0.249682 0.240099 1.040
## race_ethnicity_cleanBlack Or African American 0.019617 0.087728 0.224
## race_ethnicity_cleanHispanic or Latino -0.124437 0.080182 -1.552
## race_ethnicity_cleanWhite Not Hispanic -0.164024 0.064764 -2.533
## race_ethnicity_cleanOther/Unknown/Declined -0.083115 0.081160 -1.024
## clean_sexMale -0.083589 0.039503 -2.116
## clean_sexNon-Binary 0.064060 0.168219 0.381
## Pr(>|t|)
## (Intercept) 0.1089
## FW_stridetimesecmean <2e-16 ***
## demoEHR_Age 0.0129 *
## demoEHR_DiseaseDuration 0.1603
## ms_dx_condensedProgressive MS 0.1929
## ms_dx_condensedMS, Subtype Not Specified 0.2996
## race_ethnicity_cleanBlack Or African American 0.8233
## race_ethnicity_cleanHispanic or Latino 0.1222
## race_ethnicity_cleanWhite Not Hispanic 0.0121 *
## race_ethnicity_cleanOther/Unknown/Declined 0.3070
## clean_sexMale 0.0356 *
## clean_sexNon-Binary 0.7037
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2337 on 202 degrees of freedom
## Multiple R-squared: 0.739, Adjusted R-squared: 0.7248
## F-statistic: 51.99 on 11 and 202 DF, p-value: < 2.2e-16
# video unique IDs
metric_regression(zeno_fw_uniqueid_df, t25fw_log, FW_stridetimesecmean)## [1] "Data Frame: zeno_fw_uniqueid_df"
## [1] "t25fw_log ~ FW_stridetimesecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.54712 -0.16589 -0.01593 0.15256 0.81779
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.20066 0.09182 2.185 0.0304 *
## FW_stridetimesecmean 1.42891 0.08830 16.182 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2604 on 150 degrees of freedom
## Multiple R-squared: 0.6358, Adjusted R-squared: 0.6334
## F-statistic: 261.9 on 1 and 150 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ FW_stridetimesecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.50379 -0.13670 -0.01517 0.14663 0.82463
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.162934 0.136741 1.192
## FW_stridetimesecmean 1.339424 0.094838 14.123
## demoEHR_Age 0.004710 0.001989 2.367
## demoEHR_DiseaseDuration 0.004681 0.002972 1.575
## ms_dx_condensedProgressive MS 0.094059 0.062420 1.507
## ms_dx_condensedMS, Subtype Not Specified -0.179475 0.177031 -1.014
## race_ethnicity_cleanBlack Or African American -0.033107 0.104768 -0.316
## race_ethnicity_cleanHispanic or Latino -0.164976 0.097217 -1.697
## race_ethnicity_cleanWhite Not Hispanic -0.180499 0.079757 -2.263
## race_ethnicity_cleanOther/Unknown/Declined -0.082786 0.102405 -0.808
## clean_sexMale -0.075163 0.047209 -1.592
## clean_sexNon-Binary 0.013783 0.245312 0.056
## Pr(>|t|)
## (Intercept) 0.2355
## FW_stridetimesecmean <2e-16 ***
## demoEHR_Age 0.0193 *
## demoEHR_DiseaseDuration 0.1174
## ms_dx_condensedProgressive MS 0.1341
## ms_dx_condensedMS, Subtype Not Specified 0.3124
## race_ethnicity_cleanBlack Or African American 0.7525
## race_ethnicity_cleanHispanic or Latino 0.0919 .
## race_ethnicity_cleanWhite Not Hispanic 0.0252 *
## race_ethnicity_cleanOther/Unknown/Declined 0.4202
## clean_sexMale 0.1136
## clean_sexNon-Binary 0.9553
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2422 on 140 degrees of freedom
## Multiple R-squared: 0.706, Adjusted R-squared: 0.6829
## F-statistic: 30.57 on 11 and 140 DF, p-value: < 2.2e-16
Univariate - FW, cadence–> T25fw
#FWS
sum(is.finite(zeno_fw_df$mean_cadence_step_per_min_pose_zv))## [1] 170# cadence model
metric_regression(zeno_fw_df, t25fw_log, mean_cadence_step_per_min_pose_zv)## [1] "Data Frame: zeno_fw_df"## Warning: Removed 44 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.07674 -0.20136 -0.04297 0.16803 1.88356
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.998170 0.163617 18.324 < 2e-16 ***
## mean_cadence_step_per_min_pose_zv -0.011313 0.001342 -8.431 1.49e-14 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3908 on 168 degrees of freedom
## (44 observations deleted due to missingness)
## Multiple R-squared: 0.2973, Adjusted R-squared: 0.2931
## F-statistic: 71.09 on 1 and 168 DF, p-value: 1.491e-14
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.84863 -0.19022 -0.03352 0.14524 1.55267
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.412344 0.207150 11.645
## mean_cadence_step_per_min_pose_zv -0.008895 0.001279 -6.956
## demoEHR_Age 0.007766 0.002838 2.736
## demoEHR_DiseaseDuration -0.001864 0.003841 -0.485
## ms_dx_condensedProgressive MS 0.362424 0.084484 4.290
## ms_dx_condensedMS, Subtype Not Specified 0.365477 0.357713 1.022
## race_ethnicity_cleanBlack Or African American 0.260124 0.140827 1.847
## race_ethnicity_cleanHispanic or Latino -0.030042 0.129058 -0.233
## race_ethnicity_cleanWhite Not Hispanic -0.163962 0.106890 -1.534
## race_ethnicity_cleanOther/Unknown/Declined -0.076245 0.140441 -0.543
## clean_sexMale -0.125360 0.064172 -1.953
## clean_sexNon-Binary 0.101057 0.251071 0.403
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## mean_cadence_step_per_min_pose_zv 8.74e-11 ***
## demoEHR_Age 0.00693 **
## demoEHR_DiseaseDuration 0.62812
## ms_dx_condensedProgressive MS 3.10e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.30848
## race_ethnicity_cleanBlack Or African American 0.06660 .
## race_ethnicity_cleanHispanic or Latino 0.81623
## race_ethnicity_cleanWhite Not Hispanic 0.12705
## race_ethnicity_cleanOther/Unknown/Declined 0.58797
## clean_sexMale 0.05253 .
## clean_sexNon-Binary 0.68786
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3462 on 158 degrees of freedom
## (44 observations deleted due to missingness)
## Multiple R-squared: 0.4815, Adjusted R-squared: 0.4454
## F-statistic: 13.34 on 11 and 158 DF, p-value: < 2.2e-16
# ground truth cadence from mat
metric_regression(zeno_fw_df, t25fw_log, FW_cadencestepsminmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ FW_cadencestepsminmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.60589 -0.17693 -0.03551 0.14525 1.39089
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.3846730 0.1149938 29.43 <2e-16 ***
## FW_cadencestepsminmean -0.0141112 0.0009113 -15.48 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3058 on 212 degrees of freedom
## Multiple R-squared: 0.5307, Adjusted R-squared: 0.5285
## F-statistic: 239.8 on 1 and 212 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ FW_cadencestepsminmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.65963 -0.19051 -0.02547 0.14472 1.21522
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 3.0470306 0.1715600 17.761
## FW_cadencestepsminmean -0.0124672 0.0009762 -12.771
## demoEHR_Age 0.0037984 0.0021767 1.745
## demoEHR_DiseaseDuration 0.0025454 0.0029013 0.877
## ms_dx_condensedProgressive MS 0.1796295 0.0622120 2.887
## ms_dx_condensedMS, Subtype Not Specified 0.3143551 0.2967606 1.059
## race_ethnicity_cleanBlack Or African American 0.0692572 0.1082362 0.640
## race_ethnicity_cleanHispanic or Latino -0.0507755 0.0989808 -0.513
## race_ethnicity_cleanWhite Not Hispanic -0.1200663 0.0800796 -1.499
## race_ethnicity_cleanOther/Unknown/Declined -0.0657912 0.1002991 -0.656
## clean_sexMale -0.1084007 0.0489306 -2.215
## clean_sexNon-Binary 0.0108231 0.2078188 0.052
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## FW_cadencestepsminmean < 2e-16 ***
## demoEHR_Age 0.08250 .
## demoEHR_DiseaseDuration 0.38136
## ms_dx_condensedProgressive MS 0.00431 **
## ms_dx_condensedMS, Subtype Not Specified 0.29073
## race_ethnicity_cleanBlack Or African American 0.52298
## race_ethnicity_cleanHispanic or Latino 0.60852
## race_ethnicity_cleanWhite Not Hispanic 0.13535
## race_ethnicity_cleanOther/Unknown/Declined 0.51260
## clean_sexMale 0.02785 *
## clean_sexNon-Binary 0.95852
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2887 on 202 degrees of freedom
## Multiple R-squared: 0.6015, Adjusted R-squared: 0.5798
## F-statistic: 27.72 on 11 and 202 DF, p-value: < 2.2e-16
# unique IDs
metric_regression(zeno_fw_uniqueid_df, t25fw_log, mean_cadence_step_per_min_pose_zv)## [1] "Data Frame: zeno_fw_uniqueid_df"## Warning: Removed 38 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9717 -0.2306 -0.0546 0.1739 1.6470
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.842617 0.196824 14.442 < 2e-16 ***
## mean_cadence_step_per_min_pose_zv -0.010079 0.001656 -6.086 1.65e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3976 on 112 degrees of freedom
## (38 observations deleted due to missingness)
## Multiple R-squared: 0.2485, Adjusted R-squared: 0.2418
## F-statistic: 37.04 on 1 and 112 DF, p-value: 1.654e-08
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.72354 -0.21376 -0.03584 0.15569 1.35513
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.281242 0.258885 8.812
## mean_cadence_step_per_min_pose_zv -0.007361 0.001634 -4.505
## demoEHR_Age 0.007930 0.003230 2.455
## demoEHR_DiseaseDuration -0.001300 0.004870 -0.267
## ms_dx_condensedProgressive MS 0.398632 0.104676 3.808
## ms_dx_condensedMS, Subtype Not Specified 0.023236 0.261059 0.089
## race_ethnicity_cleanBlack Or African American 0.113740 0.173867 0.654
## race_ethnicity_cleanHispanic or Latino -0.071318 0.163091 -0.437
## race_ethnicity_cleanWhite Not Hispanic -0.262139 0.136289 -1.923
## race_ethnicity_cleanOther/Unknown/Declined -0.151719 0.186778 -0.812
## clean_sexMale -0.087965 0.077644 -1.133
## clean_sexNon-Binary 0.051987 0.361983 0.144
## Pr(>|t|)
## (Intercept) 3.44e-14 ***
## mean_cadence_step_per_min_pose_zv 1.77e-05 ***
## demoEHR_Age 0.015760 *
## demoEHR_DiseaseDuration 0.789963
## ms_dx_condensedProgressive MS 0.000239 ***
## ms_dx_condensedMS, Subtype Not Specified 0.929251
## race_ethnicity_cleanBlack Or African American 0.514472
## race_ethnicity_cleanHispanic or Latino 0.662825
## race_ethnicity_cleanWhite Not Hispanic 0.057218 .
## race_ethnicity_cleanOther/Unknown/Declined 0.418513
## clean_sexMale 0.259900
## clean_sexNon-Binary 0.886087
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3553 on 102 degrees of freedom
## (38 observations deleted due to missingness)
## Multiple R-squared: 0.4537, Adjusted R-squared: 0.3948
## F-statistic: 7.7 on 11 and 102 DF, p-value: 1.51e-09
Univariate - FW, stride width–> T25fw
# Fast Walking Speed
sum(is.finite(zeno_fw_df$stride_width_mean_cm_pose_zv))## [1] 169## model
metric_regression(zeno_fw_df, t25fw_log, stride_width_median_cm_pose_zv)## [1] "Data Frame: zeno_fw_df"## Warning: Removed 45 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.68230 -0.24335 -0.09593 0.09710 2.33305
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.27803 0.15463 8.265 4.14e-14 ***
## stride_width_median_cm_pose_zv 0.03012 0.01254 2.403 0.0174 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4588 on 167 degrees of freedom
## (45 observations deleted due to missingness)
## Multiple R-squared: 0.03341, Adjusted R-squared: 0.02762
## F-statistic: 5.772 on 1 and 167 DF, p-value: 0.01738
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.73201 -0.21750 -0.04661 0.12130 1.71544
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.188740 0.224546 5.294
## stride_width_median_cm_pose_zv 0.014262 0.011550 1.235
## demoEHR_Age 0.008147 0.003227 2.525
## demoEHR_DiseaseDuration -0.004876 0.004376 -1.114
## ms_dx_condensedProgressive MS 0.538666 0.091816 5.867
## ms_dx_condensedMS, Subtype Not Specified 0.236494 0.406830 0.581
## race_ethnicity_cleanBlack Or African American 0.267262 0.160062 1.670
## race_ethnicity_cleanHispanic or Latino -0.091687 0.148082 -0.619
## race_ethnicity_cleanWhite Not Hispanic -0.194807 0.123103 -1.582
## race_ethnicity_cleanOther/Unknown/Declined -0.096982 0.162721 -0.596
## clean_sexMale -0.128876 0.073409 -1.756
## clean_sexNon-Binary -0.015069 0.285596 -0.053
## Pr(>|t|)
## (Intercept) 3.97e-07 ***
## stride_width_median_cm_pose_zv 0.2187
## demoEHR_Age 0.0126 *
## demoEHR_DiseaseDuration 0.2669
## ms_dx_condensedProgressive MS 2.55e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.5619
## race_ethnicity_cleanBlack Or African American 0.0970 .
## race_ethnicity_cleanHispanic or Latino 0.5367
## race_ethnicity_cleanWhite Not Hispanic 0.1156
## race_ethnicity_cleanOther/Unknown/Declined 0.5520
## clean_sexMale 0.0811 .
## clean_sexNon-Binary 0.9580
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3933 on 157 degrees of freedom
## (45 observations deleted due to missingness)
## Multiple R-squared: 0.3323, Adjusted R-squared: 0.2855
## F-statistic: 7.103 on 11 and 157 DF, p-value: 1.006e-09
## pressure mat stride width model
metric_regression(zeno_fw_df, t25fw_log, FW_stridewidthcmmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ FW_stridewidthcmmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.6943 -0.2283 -0.1038 0.1228 2.1932
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.20143 0.07999 15.02 < 2e-16 ***
## FW_stridewidthcmmean 0.04381 0.00758 5.78 2.65e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4149 on 212 degrees of freedom
## Multiple R-squared: 0.1361, Adjusted R-squared: 0.132
## F-statistic: 33.4 on 1 and 212 DF, p-value: 2.651e-08
## [1] "t25fw_log ~ FW_stridewidthcmmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.86562 -0.21225 -0.05185 0.14417 1.71516
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.903279 0.159311 5.670
## FW_stridewidthcmmean 0.036545 0.007050 5.184
## demoEHR_Age 0.008181 0.002723 3.004
## demoEHR_DiseaseDuration -0.004739 0.003594 -1.319
## ms_dx_condensedProgressive MS 0.411824 0.074046 5.562
## ms_dx_condensedMS, Subtype Not Specified 0.312724 0.375138 0.834
## race_ethnicity_cleanBlack Or African American 0.137154 0.136523 1.005
## race_ethnicity_cleanHispanic or Latino -0.007251 0.125396 -0.058
## race_ethnicity_cleanWhite Not Hispanic -0.074280 0.102369 -0.726
## race_ethnicity_cleanOther/Unknown/Declined -0.011317 0.127545 -0.089
## clean_sexMale -0.089020 0.061863 -1.439
## clean_sexNon-Binary 0.020490 0.262506 0.078
## Pr(>|t|)
## (Intercept) 4.89e-08 ***
## FW_stridewidthcmmean 5.26e-07 ***
## demoEHR_Age 0.003 **
## demoEHR_DiseaseDuration 0.189
## ms_dx_condensedProgressive MS 8.40e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.405
## race_ethnicity_cleanBlack Or African American 0.316
## race_ethnicity_cleanHispanic or Latino 0.954
## race_ethnicity_cleanWhite Not Hispanic 0.469
## race_ethnicity_cleanOther/Unknown/Declined 0.929
## clean_sexMale 0.152
## clean_sexNon-Binary 0.938
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3647 on 202 degrees of freedom
## Multiple R-squared: 0.3642, Adjusted R-squared: 0.3296
## F-statistic: 10.52 on 11 and 202 DF, p-value: 3.516e-15
## unique IDs
metric_regression(zeno_fw_uniqueid_df, t25fw_log, stride_width_median_cm_pose_zv)## [1] "Data Frame: zeno_fw_uniqueid_df"## Warning: Removed 38 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.5568 -0.2573 -0.0947 0.1019 2.1141
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.47875 0.17833 8.292 2.75e-13 ***
## stride_width_median_cm_pose_zv 0.01559 0.01439 1.083 0.281
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4563 on 112 degrees of freedom
## (38 observations deleted due to missingness)
## Multiple R-squared: 0.01037, Adjusted R-squared: 0.001535
## F-statistic: 1.174 on 1 and 112 DF, p-value: 0.281
## [1] "t25fw_log ~ stride_width_median_cm_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.90254 -0.22208 -0.03822 0.13187 1.50903
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.4609357 0.2642588 5.528
## stride_width_median_cm_pose_zv -0.0008094 0.0132444 -0.061
## demoEHR_Age 0.0069604 0.0035288 1.972
## demoEHR_DiseaseDuration -0.0013514 0.0053694 -0.252
## ms_dx_condensedProgressive MS 0.5878604 0.1062883 5.531
## ms_dx_condensedMS, Subtype Not Specified -0.0330026 0.2878626 -0.115
## race_ethnicity_cleanBlack Or African American 0.1380764 0.1903464 0.725
## race_ethnicity_cleanHispanic or Latino -0.1339586 0.1817699 -0.737
## race_ethnicity_cleanWhite Not Hispanic -0.2676614 0.1517299 -1.764
## race_ethnicity_cleanOther/Unknown/Declined -0.1772072 0.2075663 -0.854
## clean_sexMale -0.0783511 0.0858632 -0.913
## clean_sexNon-Binary 0.0018673 0.3981398 0.005
## Pr(>|t|)
## (Intercept) 2.50e-07 ***
## stride_width_median_cm_pose_zv 0.9514
## demoEHR_Age 0.0513 .
## demoEHR_DiseaseDuration 0.8018
## ms_dx_condensedProgressive MS 2.47e-07 ***
## ms_dx_condensedMS, Subtype Not Specified 0.9090
## race_ethnicity_cleanBlack Or African American 0.4699
## race_ethnicity_cleanHispanic or Latino 0.4628
## race_ethnicity_cleanWhite Not Hispanic 0.0807 .
## race_ethnicity_cleanOther/Unknown/Declined 0.3953
## clean_sexMale 0.3637
## clean_sexNon-Binary 0.9963
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.389 on 102 degrees of freedom
## (38 observations deleted due to missingness)
## Multiple R-squared: 0.345, Adjusted R-squared: 0.2744
## F-statistic: 4.884 on 11 and 102 DF, p-value: 4.897e-06
Univariate - FW, Stance Time –> T25fw
Stance/swing/double/single support measures not calculated for all participants
# FW
sum(is.finite(zeno_fw_df$foot1_stance_time_mean_pose_zv))## [1] 31#video model
metric_regression(zeno_fw_df, t25fw_log, foot1_stance_time_mean_pose_zv)## [1] "Data Frame: zeno_fw_df"## Warning: Removed 183 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.54928 -0.17667 -0.03690 0.07738 0.91980
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.9376 0.2347 3.995 0.000406 ***
## foot1_stance_time_mean_pose_zv 1.0468 0.3311 3.161 0.003663 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3083 on 29 degrees of freedom
## (183 observations deleted due to missingness)
## Multiple R-squared: 0.2563, Adjusted R-squared: 0.2306
## F-statistic: 9.993 on 1 and 29 DF, p-value: 0.003663
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.33303 -0.13277 -0.03469 0.12341 0.73569
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.476510 0.391903 3.768
## foot1_stance_time_mean_pose_zv 0.516747 0.358040 1.443
## demoEHR_Age -0.010695 0.008343 -1.282
## demoEHR_DiseaseDuration 0.014240 0.014096 1.010
## ms_dx_condensedProgressive MS 0.509076 0.222935 2.284
## race_ethnicity_cleanBlack Or African American 0.624933 0.303581 2.059
## race_ethnicity_cleanHispanic or Latino 0.190468 0.209485 0.909
## race_ethnicity_cleanWhite Not Hispanic 0.066883 0.196828 0.340
## race_ethnicity_cleanOther/Unknown/Declined 0.539383 0.245188 2.200
## clean_sexMale -0.065235 0.150037 -0.435
## Pr(>|t|)
## (Intercept) 0.00113 **
## foot1_stance_time_mean_pose_zv 0.16370
## demoEHR_Age 0.21382
## demoEHR_DiseaseDuration 0.32386
## ms_dx_condensedProgressive MS 0.03292 *
## race_ethnicity_cleanBlack Or African American 0.05216 .
## race_ethnicity_cleanHispanic or Latino 0.37355
## race_ethnicity_cleanWhite Not Hispanic 0.73738
## race_ethnicity_cleanOther/Unknown/Declined 0.03915 *
## clean_sexMale 0.66815
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2663 on 21 degrees of freedom
## (183 observations deleted due to missingness)
## Multiple R-squared: 0.5981, Adjusted R-squared: 0.4258
## F-statistic: 3.472 on 9 and 21 DF, p-value: 0.008969
# no pressure mat value
# unique IDs
metric_regression(zeno_fw_uniqueid_df, t25fw_log, foot1_stance_time_mean_pose_zv)## [1] "Data Frame: zeno_fw_uniqueid_df"## Warning: Removed 135 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.45364 -0.15070 -0.04724 0.08034 0.87527
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.8700 0.3596 2.420 0.0287 *
## foot1_stance_time_mean_pose_zv 1.1929 0.4799 2.485 0.0252 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3462 on 15 degrees of freedom
## (135 observations deleted due to missingness)
## Multiple R-squared: 0.2917, Adjusted R-squared: 0.2445
## F-statistic: 6.178 on 1 and 15 DF, p-value: 0.02522
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.33986 -0.08339 0.00000 0.07865 0.51272
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.726126 0.750183 2.301
## foot1_stance_time_mean_pose_zv -0.107262 0.737599 -0.145
## demoEHR_Age -0.001685 0.013411 -0.126
## demoEHR_DiseaseDuration -0.004882 0.023913 -0.204
## ms_dx_condensedProgressive MS 0.616849 0.345036 1.788
## race_ethnicity_cleanBlack Or African American 1.325559 0.525697 2.522
## race_ethnicity_cleanHispanic or Latino 0.326011 0.350369 0.930
## race_ethnicity_cleanWhite Not Hispanic 0.033550 0.345840 0.097
## race_ethnicity_cleanOther/Unknown/Declined 0.143184 0.452789 0.316
## clean_sexMale -0.149444 0.185316 -0.806
## Pr(>|t|)
## (Intercept) 0.0549 .
## foot1_stance_time_mean_pose_zv 0.8885
## demoEHR_Age 0.9036
## demoEHR_DiseaseDuration 0.8440
## ms_dx_condensedProgressive MS 0.1170
## race_ethnicity_cleanBlack Or African American 0.0397 *
## race_ethnicity_cleanHispanic or Latino 0.3831
## race_ethnicity_cleanWhite Not Hispanic 0.9254
## race_ethnicity_cleanOther/Unknown/Declined 0.7611
## clean_sexMale 0.4465
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3045 on 7 degrees of freedom
## (135 observations deleted due to missingness)
## Multiple R-squared: 0.7443, Adjusted R-squared: 0.4156
## F-statistic: 2.264 on 9 and 7 DF, p-value: 0.147
Univariate - FW, SWing Time –> T25fw
# FW
sum(is.finite(zeno_fw_df$foot1_swing_time_mean_pose_zv))## [1] 31ggplot(data = zeno_fw_df, aes(x = foot1_swing_time_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 183 rows containing missing values (`geom_point()`).
ggplot(data = zeno_fw_df, aes(x = foot1_swing_per_mean_pose_zv, y = t25fw_log)) +
geom_point()## Warning: Removed 183 rows containing missing values (`geom_point()`).
FW - remove outlier?? %, shouldn’t be greater than 100…
### Univariate - FW, Single and Double Support –> T25fw
# Fast Walking
sum(is.finite(zeno_fw_df$foot1_term_double_support_time_mean_pose_zv))## [1] 31# try terminal double support time
metric_regression(zeno_fw_df, t25fw_log, foot1_term_double_support_time_mean_pose_zv)## [1] "Data Frame: zeno_fw_df"## Warning: Removed 183 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.48458 -0.12730 -0.03923 0.09861 0.68192
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.3462 0.0804 16.74
## foot1_term_double_support_time_mean_pose_zv 1.9114 0.3957 4.83
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## foot1_term_double_support_time_mean_pose_zv 4.06e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2661 on 29 degrees of freedom
## (183 observations deleted due to missingness)
## Multiple R-squared: 0.4458, Adjusted R-squared: 0.4267
## F-statistic: 23.33 on 1 and 29 DF, p-value: 4.065e-05
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.35764 -0.11927 -0.02492 0.10910 0.59988
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.527598 0.267627 5.708
## foot1_term_double_support_time_mean_pose_zv 1.360813 0.476985 2.853
## demoEHR_Age -0.008067 0.007507 -1.074
## demoEHR_DiseaseDuration 0.010679 0.012634 0.845
## ms_dx_condensedProgressive MS 0.355741 0.205824 1.728
## race_ethnicity_cleanBlack Or African American 0.472928 0.277439 1.705
## race_ethnicity_cleanHispanic or Latino 0.213172 0.184110 1.158
## race_ethnicity_cleanWhite Not Hispanic 0.073562 0.173978 0.423
## race_ethnicity_cleanOther/Unknown/Declined 0.576028 0.218468 2.637
## clean_sexMale 0.044534 0.137062 0.325
## Pr(>|t|)
## (Intercept) 1.15e-05 ***
## foot1_term_double_support_time_mean_pose_zv 0.00953 **
## demoEHR_Age 0.29480
## demoEHR_DiseaseDuration 0.40746
## ms_dx_condensedProgressive MS 0.09860 .
## race_ethnicity_cleanBlack Or African American 0.10302
## race_ethnicity_cleanHispanic or Latino 0.25992
## race_ethnicity_cleanWhite Not Hispanic 0.67672
## race_ethnicity_cleanOther/Unknown/Declined 0.01542 *
## clean_sexMale 0.74846
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2371 on 21 degrees of freedom
## (183 observations deleted due to missingness)
## Multiple R-squared: 0.6816, Adjusted R-squared: 0.5452
## F-statistic: 4.995 on 9 and 21 DF, p-value: 0.00115
# unique IDs
metric_regression(zeno_pws_uniqueid_df, t25fw_log, foot1_term_double_support_time_mean_pose_zv)## [1] "Data Frame: zeno_pws_uniqueid_df"## Warning: Removed 122 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47957 -0.21192 -0.01916 0.19843 0.75915
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.48305 0.08981 16.513
## foot1_term_double_support_time_mean_pose_zv 1.04515 0.32959 3.171
## Pr(>|t|)
## (Intercept) 5.77e-16 ***
## foot1_term_double_support_time_mean_pose_zv 0.00366 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2774 on 28 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.2642, Adjusted R-squared: 0.238
## F-statistic: 10.06 on 1 and 28 DF, p-value: 0.003664
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.29714 -0.19772 -0.03331 0.10630 0.63890
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.6517784 0.3007771 5.492
## foot1_term_double_support_time_mean_pose_zv 0.8110046 0.4014502 2.020
## demoEHR_Age -0.0007846 0.0051959 -0.151
## demoEHR_DiseaseDuration 0.0059087 0.0074536 0.793
## ms_dx_condensedProgressive MS 0.1760955 0.1774567 0.992
## race_ethnicity_cleanBlack Or African American 0.2183771 0.2828617 0.772
## race_ethnicity_cleanHispanic or Latino -0.1146077 0.2180385 -0.526
## race_ethnicity_cleanWhite Not Hispanic -0.2338653 0.1967099 -1.189
## clean_sexMale -0.0104773 0.1336019 -0.078
## Pr(>|t|)
## (Intercept) 1.9e-05 ***
## foot1_term_double_support_time_mean_pose_zv 0.0563 .
## demoEHR_Age 0.8814
## demoEHR_DiseaseDuration 0.4368
## ms_dx_condensedProgressive MS 0.3323
## race_ethnicity_cleanBlack Or African American 0.4487
## race_ethnicity_cleanHispanic or Latino 0.6047
## race_ethnicity_cleanWhite Not Hispanic 0.2478
## clean_sexMale 0.9382
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2816 on 21 degrees of freedom
## (122 observations deleted due to missingness)
## Multiple R-squared: 0.4315, Adjusted R-squared: 0.2149
## F-statistic: 1.992 on 8 and 21 DF, p-value: 0.09835
# mat double support
metric_regression(zeno_fw_df, t25fw_log, FW_totaldsupportmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ FW_totaldsupportmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.75092 -0.26545 -0.10031 0.09632 2.28499
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.626e+00 2.973e-02 54.698 < 2e-16 ***
## FW_totaldsupportmean 1.449e-05 4.085e-06 3.547 0.00048 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4337 on 212 degrees of freedom
## Multiple R-squared: 0.05601, Adjusted R-squared: 0.05155
## F-statistic: 12.58 on 1 and 212 DF, p-value: 0.0004804
## [1] "t25fw_log ~ FW_totaldsupportmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.01867 -0.21125 -0.05201 0.12558 1.81766
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.354e+00 1.418e-01 9.549
## FW_totaldsupportmean 1.002e-05 3.659e-06 2.739
## demoEHR_Age 7.176e-03 2.852e-03 2.516
## demoEHR_DiseaseDuration -3.580e-03 3.778e-03 -0.947
## ms_dx_condensedProgressive MS 4.391e-01 7.723e-02 5.686
## ms_dx_condensedMS, Subtype Not Specified 2.048e-01 3.917e-01 0.523
## race_ethnicity_cleanBlack Or African American 1.847e-01 1.424e-01 1.297
## race_ethnicity_cleanHispanic or Latino -6.004e-02 1.307e-01 -0.459
## race_ethnicity_cleanWhite Not Hispanic -1.622e-01 1.057e-01 -1.535
## race_ethnicity_cleanOther/Unknown/Declined -9.178e-02 1.324e-01 -0.693
## clean_sexMale -5.459e-02 6.445e-02 -0.847
## clean_sexNon-Binary 1.435e-02 2.744e-01 0.052
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## FW_totaldsupportmean 0.00672 **
## demoEHR_Age 0.01263 *
## demoEHR_DiseaseDuration 0.34452
## ms_dx_condensedProgressive MS 4.51e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.60159
## race_ethnicity_cleanBlack Or African American 0.19620
## race_ethnicity_cleanHispanic or Latino 0.64639
## race_ethnicity_cleanWhite Not Hispanic 0.12627
## race_ethnicity_cleanOther/Unknown/Declined 0.48894
## clean_sexMale 0.39799
## clean_sexNon-Binary 0.95833
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3811 on 202 degrees of freedom
## Multiple R-squared: 0.3055, Adjusted R-squared: 0.2676
## F-statistic: 8.076 on 11 and 202 DF, p-value: 1.235e-11
zeno_fw_df$log_FW_totaldsupportmean <- log(zeno_fw_df$FW_totaldsupportmean)
metric_regression(zeno_fw_df, t25fw_log, log_FW_totaldsupportmean)## [1] "Data Frame: zeno_fw_df"
## [1] "t25fw_log ~ log_FW_totaldsupportmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.81592 -0.20781 -0.05199 0.10341 1.91655
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.2990 0.1265 2.364 0.019 *
## log_FW_totaldsupportmean 0.4034 0.0375 10.757 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3591 on 212 degrees of freedom
## Multiple R-squared: 0.3531, Adjusted R-squared: 0.35
## F-statistic: 115.7 on 1 and 212 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ log_FW_totaldsupportmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.41703 -0.18370 -0.05231 0.13717 1.68433
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.4454946 0.1636201 2.723
## log_FW_totaldsupportmean 0.3176662 0.0379110 8.379
## demoEHR_Age 0.0039960 0.0025350 1.576
## demoEHR_DiseaseDuration -0.0008363 0.0033274 -0.251
## ms_dx_condensedProgressive MS 0.3163545 0.0695423 4.549
## ms_dx_condensedMS, Subtype Not Specified 0.1235777 0.3437682 0.359
## race_ethnicity_cleanBlack Or African American 0.1344477 0.1250645 1.075
## race_ethnicity_cleanHispanic or Latino -0.0983261 0.1147326 -0.857
## race_ethnicity_cleanWhite Not Hispanic -0.1336072 0.0927248 -1.441
## race_ethnicity_cleanOther/Unknown/Declined -0.0841658 0.1161419 -0.725
## clean_sexMale -0.0459628 0.0565304 -0.813
## clean_sexNon-Binary 0.0438553 0.2407309 0.182
## Pr(>|t|)
## (Intercept) 0.00704 **
## log_FW_totaldsupportmean 8.95e-15 ***
## demoEHR_Age 0.11651
## demoEHR_DiseaseDuration 0.80180
## ms_dx_condensedProgressive MS 9.28e-06 ***
## ms_dx_condensedMS, Subtype Not Specified 0.71961
## race_ethnicity_cleanBlack Or African American 0.28365
## race_ethnicity_cleanHispanic or Latino 0.39246
## race_ethnicity_cleanWhite Not Hispanic 0.15116
## race_ethnicity_cleanOther/Unknown/Declined 0.46949
## clean_sexMale 0.41714
## clean_sexNon-Binary 0.85563
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3344 on 202 degrees of freedom
## Multiple R-squared: 0.4655, Adjusted R-squared: 0.4364
## F-statistic: 15.99 on 11 and 202 DF, p-value: < 2.2e-16
Multivariate - FW, Gait metrics only –> T25fw
Metrics only - not including double support/stance measures, too many missing. May include after improving code
# FW
fw_t25fw_multivar_model <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.74282 -0.16267 -0.02336 0.11850 1.43166
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.299713 0.382281 3.400 0.000854 ***
## log_delta_pix_h_rel_median_pose_zv -0.361306 0.051372 -7.033 5.76e-11 ***
## stride_time_median_sec_pose_zv 0.516821 0.207082 2.496 0.013597 *
## mean_cadence_step_per_min_pose_zv -0.004330 0.001756 -2.467 0.014707 *
## stride_width_median_cm_pose_zv -0.001622 0.008724 -0.186 0.852740
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2884 on 158 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.5429, Adjusted R-squared: 0.5313
## F-statistic: 46.91 on 4 and 158 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model))
fw_t25fw_multivar_model_2 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv *
stride_time_median_sec_pose_zv *
mean_cadence_step_per_min_pose_zv *
stride_width_median_cm_pose_zv,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model_2)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv *
## stride_time_median_sec_pose_zv * mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.70277 -0.14274 -0.00954 0.13379 1.21479
##
## Coefficients:
## Estimate
## (Intercept) -13.741585
## log_delta_pix_h_rel_median_pose_zv -5.900066
## stride_time_median_sec_pose_zv 17.362875
## mean_cadence_step_per_min_pose_zv 0.113749
## stride_width_median_cm_pose_zv 1.245241
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 3.941885
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.028876
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.132176
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.462095
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -1.540419
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.010832
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.010176
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -0.479686
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.003478
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.013381
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.003381
## Std. Error
## (Intercept) 6.782085
## log_delta_pix_h_rel_median_pose_zv 3.843556
## stride_time_median_sec_pose_zv 7.082843
## mean_cadence_step_per_min_pose_zv 0.051170
## stride_width_median_cm_pose_zv 0.589806
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 3.839776
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.035562
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.057648
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.327849
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.611170
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.004566
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.037112
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.319171
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.002879
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.005086
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.002929
## t value
## (Intercept) -2.026
## log_delta_pix_h_rel_median_pose_zv -1.535
## stride_time_median_sec_pose_zv 2.451
## mean_cadence_step_per_min_pose_zv 2.223
## stride_width_median_cm_pose_zv 2.111
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 1.027
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.812
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -2.293
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 1.409
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -2.520
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -2.372
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv -0.274
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv -1.503
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -1.208
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 2.631
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 1.154
## Pr(>|t|)
## (Intercept) 0.04456
## log_delta_pix_h_rel_median_pose_zv 0.12692
## stride_time_median_sec_pose_zv 0.01540
## mean_cadence_step_per_min_pose_zv 0.02774
## stride_width_median_cm_pose_zv 0.03644
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 0.30630
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.41811
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.02328
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.16081
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.01279
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.01898
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.78432
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.13501
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.22890
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.00943
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.25025
##
## (Intercept) *
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv *
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv **
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2594 on 147 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.6561, Adjusted R-squared: 0.621
## F-statistic: 18.69 on 15 and 147 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model_2))
Multivariate - FW, Gait metrics + disease + demographics –> T25fw
# FW
# add MS subtype
fw_t25fw_multivar_model_3 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model_3)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.72744 -0.16550 -0.01536 0.11962 1.30768
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.353083 0.378232 3.577 0.000463
## log_delta_pix_h_rel_median_pose_zv -0.348079 0.051143 -6.806 2.03e-10
## stride_time_median_sec_pose_zv 0.445776 0.207016 2.153 0.032828
## mean_cadence_step_per_min_pose_zv -0.004189 0.001739 -2.409 0.017166
## stride_width_median_cm_pose_zv -0.002735 0.008655 -0.316 0.752408
## ms_dx_condensedProgressive MS 0.153806 0.069690 2.207 0.028776
## ms_dx_condensedMS, Subtype Not Specified 0.318720 0.287672 1.108 0.269598
##
## (Intercept) ***
## log_delta_pix_h_rel_median_pose_zv ***
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2848 on 156 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.5599, Adjusted R-squared: 0.543
## F-statistic: 33.08 on 6 and 156 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model_3))
# add age
fw_t25fw_multivar_model_4 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model_4)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age,
## data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.71450 -0.15767 -0.01378 0.14518 1.26586
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.090498 0.384480 2.836 0.00517
## log_delta_pix_h_rel_median_pose_zv -0.341129 0.050271 -6.786 2.3e-10
## stride_time_median_sec_pose_zv 0.480248 0.203628 2.358 0.01960
## mean_cadence_step_per_min_pose_zv -0.004205 0.001707 -2.464 0.01485
## stride_width_median_cm_pose_zv -0.002871 0.008496 -0.338 0.73585
## ms_dx_condensedProgressive MS 0.091722 0.072372 1.267 0.20692
## ms_dx_condensedMS, Subtype Not Specified 0.226429 0.284551 0.796 0.42740
## demoEHR_Age 0.005011 0.001907 2.628 0.00945
##
## (Intercept) **
## log_delta_pix_h_rel_median_pose_zv ***
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2796 on 155 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.5787, Adjusted R-squared: 0.5597
## F-statistic: 30.41 on 7 and 155 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model_4))
# add disease duration
fw_t25fw_multivar_model_5 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model_5)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.70918 -0.16482 -0.01377 0.14354 1.27584
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.0970471 0.3861850 2.841 0.00511
## log_delta_pix_h_rel_median_pose_zv -0.3400840 0.0505308 -6.730 3.15e-10
## stride_time_median_sec_pose_zv 0.4862678 0.2051473 2.370 0.01901
## mean_cadence_step_per_min_pose_zv -0.0042238 0.0017130 -2.466 0.01477
## stride_width_median_cm_pose_zv -0.0031944 0.0085843 -0.372 0.71032
## ms_dx_condensedProgressive MS 0.0894619 0.0729496 1.226 0.22194
## ms_dx_condensedMS, Subtype Not Specified 0.2155000 0.2875580 0.749 0.45475
## demoEHR_Age 0.0047091 0.0021461 2.194 0.02972
## demoEHR_DiseaseDuration 0.0009894 0.0031941 0.310 0.75717
##
## (Intercept) **
## log_delta_pix_h_rel_median_pose_zv ***
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age *
## demoEHR_DiseaseDuration
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2804 on 154 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.5789, Adjusted R-squared: 0.5571
## F-statistic: 26.47 on 8 and 154 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model_5))
# add race and ethnicity
fw_t25fw_multivar_model_6 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = zeno_fw_df)
summary(fw_t25fw_multivar_model_6)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = zeno_fw_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.6837 -0.1508 -0.0069 0.1173 1.2468
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.2620082 0.3798622 3.322
## log_delta_pix_h_rel_median_pose_zv -0.3244476 0.0502294 -6.459
## stride_time_median_sec_pose_zv 0.4355788 0.1990289 2.189
## mean_cadence_step_per_min_pose_zv -0.0041332 0.0016762 -2.466
## stride_width_median_cm_pose_zv -0.0095527 0.0086793 -1.101
## ms_dx_condensedProgressive MS 0.1185009 0.0715211 1.657
## ms_dx_condensedMS, Subtype Not Specified 0.2510128 0.2783538 0.902
## demoEHR_Age 0.0056524 0.0022504 2.512
## demoEHR_DiseaseDuration 0.0009945 0.0030963 0.321
## race_ethnicity_cleanBlack Or African American 0.1921663 0.1108580 1.733
## race_ethnicity_cleanHispanic or Latino -0.0770721 0.1052229 -0.732
## race_ethnicity_cleanWhite Not Hispanic -0.1256500 0.0852231 -1.474
## race_ethnicity_cleanOther/Unknown/Declined -0.1115388 0.1127321 -0.989
## Pr(>|t|)
## (Intercept) 0.00112 **
## log_delta_pix_h_rel_median_pose_zv 1.38e-09 ***
## stride_time_median_sec_pose_zv 0.03018 *
## mean_cadence_step_per_min_pose_zv 0.01480 *
## stride_width_median_cm_pose_zv 0.27282
## ms_dx_condensedProgressive MS 0.09964 .
## ms_dx_condensedMS, Subtype Not Specified 0.36862
## demoEHR_Age 0.01307 *
## demoEHR_DiseaseDuration 0.74851
## race_ethnicity_cleanBlack Or African American 0.08507 .
## race_ethnicity_cleanHispanic or Latino 0.46503
## race_ethnicity_cleanWhite Not Hispanic 0.14248
## race_ethnicity_cleanOther/Unknown/Declined 0.32405
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2712 on 150 degrees of freedom
## (51 observations deleted due to missingness)
## Multiple R-squared: 0.6163, Adjusted R-squared: 0.5856
## F-statistic: 20.07 on 12 and 150 DF, p-value: < 2.2e-16hist(resid(fw_t25fw_multivar_model_6))
Home Videos –> T25fw
Demographics and disease info
# Home Videos - all
home_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
ms_dx_condensed +
clean_sex,
data = home_df)
summary(home_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + ms_dx_condensed + clean_sex, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.60983 -0.15926 -0.04037 0.12917 1.04400
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.275621 0.179586 7.103 2.34e-09
## demoEHR_Age 0.005050 0.003670 1.376 0.1743
## demoEHR_DiseaseDuration 0.008190 0.005782 1.417 0.1621
## ms_dx_condensedProgressive MS 0.671223 0.132635 5.061 4.82e-06
## ms_dx_condensedMS, Subtype Not Specified -0.075133 0.186614 -0.403 0.6888
## clean_sexMale -0.268558 0.121983 -2.202 0.0318
## clean_sexNon-Binary 0.060679 0.235474 0.258 0.7976
##
## (Intercept) ***
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS ***
## ms_dx_condensedMS, Subtype Not Specified
## clean_sexMale *
## clean_sexNon-Binary
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3129 on 56 degrees of freedom
## Multiple R-squared: 0.4408, Adjusted R-squared: 0.3809
## F-statistic: 7.356 on 6 and 56 DF, p-value: 7.893e-06# check residuals
hist(resid(home_dem_model))
# home right -------------------------
home_r_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
ms_dx_condensed +
clean_sex,
data = home_r_df)
summary(home_r_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + ms_dx_condensed + clean_sex, data = home_r_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.4843 -0.1668 -0.0488 0.1267 1.0279
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.293913 0.269057 4.809 6.11e-05
## demoEHR_Age 0.005014 0.005572 0.900 0.37682
## demoEHR_DiseaseDuration 0.008172 0.008676 0.942 0.35528
## ms_dx_condensedProgressive MS 0.595348 0.192915 3.086 0.00491
## ms_dx_condensedMS, Subtype Not Specified -0.057804 0.282222 -0.205 0.83937
## clean_sexMale -0.334210 0.176156 -1.897 0.06941
## clean_sexNon-Binary 0.044067 0.356200 0.124 0.90253
##
## (Intercept) ***
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS **
## ms_dx_condensedMS, Subtype Not Specified
## clean_sexMale .
## clean_sexNon-Binary
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3363 on 25 degrees of freedom
## Multiple R-squared: 0.4234, Adjusted R-squared: 0.2851
## F-statistic: 3.06 on 6 and 25 DF, p-value: 0.02195# check residuals
hist(resid(home_r_dem_model))
# home left --------------------------
home_l_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
ms_dx_condensed +
clean_sex,
data = home_l_df)
summary(home_l_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + ms_dx_condensed + clean_sex, data = home_l_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.54392 -0.14622 -0.02766 0.10132 1.05961
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.229119 0.269913 4.554 0.000129
## demoEHR_Age 0.005402 0.005418 0.997 0.328671
## demoEHR_DiseaseDuration 0.008811 0.008665 1.017 0.319359
## ms_dx_condensedProgressive MS 0.794469 0.208926 3.803 0.000866
## ms_dx_condensedMS, Subtype Not Specified -0.119648 0.277337 -0.431 0.670016
## clean_sexMale -0.156355 0.193958 -0.806 0.428085
## clean_sexNon-Binary 0.091006 0.348708 0.261 0.796334
##
## (Intercept) ***
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS ***
## ms_dx_condensedMS, Subtype Not Specified
## clean_sexMale
## clean_sexNon-Binary
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3257 on 24 degrees of freedom
## Multiple R-squared: 0.4807, Adjusted R-squared: 0.3508
## F-statistic: 3.702 on 6 and 24 DF, p-value: 0.00954hist(resid(home_l_dem_model))
# home unique IDs ----------------
home_uniqueid_dem_model <- lm(t25fw_log ~ demoEHR_Age +
demoEHR_DiseaseDuration +
ms_dx_condensed +
ms_dx_condensed +
clean_sex,
data = home_uniqueid_df)
summary(home_uniqueid_dem_model)##
## Call:
## lm(formula = t25fw_log ~ demoEHR_Age + demoEHR_DiseaseDuration +
## ms_dx_condensed + ms_dx_condensed + clean_sex, data = home_uniqueid_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.49230 -0.16271 -0.02361 0.15094 0.99514
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.389490 0.303322 4.581 0.000132
## demoEHR_Age 0.002890 0.006146 0.470 0.642626
## demoEHR_DiseaseDuration 0.010047 0.009720 1.034 0.312052
## ms_dx_condensedProgressive MS 0.610104 0.201017 3.035 0.005884
## ms_dx_condensedMS, Subtype Not Specified -0.034291 0.296007 -0.116 0.908781
## clean_sexMale -0.353169 0.186145 -1.897 0.070416
## clean_sexNon-Binary 0.046192 0.373275 0.124 0.902589
##
## (Intercept) ***
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS **
## ms_dx_condensedMS, Subtype Not Specified
## clean_sexMale .
## clean_sexNon-Binary
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3505 on 23 degrees of freedom
## Multiple R-squared: 0.4144, Adjusted R-squared: 0.2617
## F-statistic: 2.713 on 6 and 23 DF, p-value: 0.03847hist(resid(home_uniqueid_dem_model))
Univariate - home, delta pixel proxy –> T25fw
# Home All
sum(is.finite(home_df$delta_pix_h_rel_median_pose_hv))## [1] 59home_df$log_delta_pix_h_rel_median_pose_hv <-log(home_df$delta_pix_h_rel_median_pose_hv)
home_df$sqrt_delta_pix_h_rel_median_pose_hv <- sqrt(home_df$delta_pix_h_rel_median_pose_hv)
#log velproxy and log t25Fw
metric_regression(home_df, t25fw_log, log_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 4 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.78400 -0.20539 -0.05828 0.16045 0.91691
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.27346 0.11925 10.679 3.21e-15 ***
## log_delta_pix_h_rel_median_pose_hv -0.28294 0.08539 -3.314 0.0016 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3549 on 57 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.1615, Adjusted R-squared: 0.1468
## F-statistic: 10.98 on 1 and 57 DF, p-value: 0.001605
## [1] "t25fw_log ~ log_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.57181 -0.09042 -0.00424 0.10968 0.44869
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.169304 0.222698 5.251
## log_delta_pix_h_rel_median_pose_hv -0.051586 0.068883 -0.749
## demoEHR_Age 0.004044 0.003202 1.263
## demoEHR_DiseaseDuration 0.001765 0.005570 0.317
## ms_dx_condensedProgressive MS 0.645632 0.108032 5.976
## ms_dx_condensedMS, Subtype Not Specified 0.015761 0.147203 0.107
## race_ethnicity_cleanBlack Or African American 1.227113 0.293983 4.174
## race_ethnicity_cleanHispanic or Latino 0.281474 0.256663 1.097
## race_ethnicity_cleanWhite Not Hispanic 0.153563 0.152929 1.004
## race_ethnicity_cleanOther/Unknown/Declined -0.019314 0.177220 -0.109
## clean_sexMale -0.274163 0.103109 -2.659
## clean_sexNon-Binary -0.001714 0.182624 -0.009
## Pr(>|t|)
## (Intercept) 3.59e-06 ***
## log_delta_pix_h_rel_median_pose_hv 0.457657
## demoEHR_Age 0.212797
## demoEHR_DiseaseDuration 0.752773
## ms_dx_condensedProgressive MS 2.92e-07 ***
## ms_dx_condensedMS, Subtype Not Specified 0.915187
## race_ethnicity_cleanBlack Or African American 0.000128 ***
## race_ethnicity_cleanHispanic or Latino 0.278373
## race_ethnicity_cleanWhite Not Hispanic 0.320450
## race_ethnicity_cleanOther/Unknown/Declined 0.913679
## clean_sexMale 0.010682 *
## clean_sexNon-Binary 0.992551
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2382 on 47 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.6884, Adjusted R-squared: 0.6155
## F-statistic: 9.44 on 11 and 47 DF, p-value: 1.178e-08
#sqrt velproxy and log t25fw
metric_regression(home_df, t25fw_log, sqrt_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 4 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.66793 -0.21030 -0.05585 0.16013 0.87504
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3898 0.2079 11.493 < 2e-16 ***
## sqrt_delta_pix_h_rel_median_pose_hv -1.3869 0.3743 -3.705 0.000479 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3479 on 57 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.1941, Adjusted R-squared: 0.18
## F-statistic: 13.73 on 1 and 57 DF, p-value: 0.0004788
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.57392 -0.09589 -0.00481 0.10940 0.44522
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.357612 0.233289 5.819
## sqrt_delta_pix_h_rel_median_pose_hv -0.212217 0.321215 -0.661
## demoEHR_Age 0.003860 0.003183 1.213
## demoEHR_DiseaseDuration 0.001979 0.005550 0.357
## ms_dx_condensedProgressive MS 0.645544 0.109676 5.886
## ms_dx_condensedMS, Subtype Not Specified 0.017943 0.147869 0.121
## race_ethnicity_cleanBlack Or African American 1.227339 0.294783 4.164
## race_ethnicity_cleanHispanic or Latino 0.280188 0.257745 1.087
## race_ethnicity_cleanWhite Not Hispanic 0.154298 0.153533 1.005
## race_ethnicity_cleanOther/Unknown/Declined -0.022284 0.177897 -0.125
## clean_sexMale -0.273926 0.104859 -2.612
## clean_sexNon-Binary -0.003771 0.182804 -0.021
## Pr(>|t|)
## (Intercept) 5.04e-07 ***
## sqrt_delta_pix_h_rel_median_pose_hv 0.512048
## demoEHR_Age 0.231237
## demoEHR_DiseaseDuration 0.722951
## ms_dx_condensedProgressive MS 4.00e-07 ***
## ms_dx_condensedMS, Subtype Not Specified 0.903937
## race_ethnicity_cleanBlack Or African American 0.000133 ***
## race_ethnicity_cleanHispanic or Latino 0.282547
## race_ethnicity_cleanWhite Not Hispanic 0.320051
## race_ethnicity_cleanOther/Unknown/Declined 0.900851
## clean_sexMale 0.012037 *
## clean_sexNon-Binary 0.983630
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2385 on 47 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.6876, Adjusted R-squared: 0.6145
## F-statistic: 9.404 on 11 and 47 DF, p-value: 1.247e-08
# benchmark - pressure mat velocity and log t25fw + demographics and disease
metric_regression(home_df, t25fw_log, PWS_velocitycmsecmean)## [1] "Data Frame: home_df"
## [1] "t25fw_log ~ PWS_velocitycmsecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.49767 -0.15691 -0.00799 0.13544 0.42438
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.868933 0.111735 25.68 <2e-16 ***
## PWS_velocitycmsecmean -0.011502 0.001024 -11.23 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2289 on 61 degrees of freedom
## Multiple R-squared: 0.6742, Adjusted R-squared: 0.6688
## F-statistic: 126.2 on 1 and 61 DF, p-value: < 2.2e-16
## [1] "t25fw_log ~ PWS_velocitycmsecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.3553 -0.1075 0.0000 0.1532 0.2638
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.2197212 0.2141056 10.367
## PWS_velocitycmsecmean -0.0075819 0.0013909 -5.451
## demoEHR_Age 0.0059091 0.0024487 2.413
## demoEHR_DiseaseDuration 0.0007665 0.0041924 0.183
## ms_dx_condensedProgressive MS 0.3399369 0.1001482 3.394
## ms_dx_condensedMS, Subtype Not Specified -0.0013667 0.1140008 -0.012
## race_ethnicity_cleanBlack Or African American 0.5810782 0.2099520 2.768
## race_ethnicity_cleanHispanic or Latino -0.1704833 0.1985851 -0.858
## race_ethnicity_cleanWhite Not Hispanic -0.1361809 0.1106516 -1.231
## race_ethnicity_cleanOther/Unknown/Declined -0.1764417 0.1238810 -1.424
## clean_sexMale -0.0219856 0.0906709 -0.242
## clean_sexNon-Binary 0.1468681 0.1454026 1.010
## Pr(>|t|)
## (Intercept) 3.72e-14 ***
## PWS_velocitycmsecmean 1.46e-06 ***
## demoEHR_Age 0.01944 *
## demoEHR_DiseaseDuration 0.85565
## ms_dx_condensedProgressive MS 0.00134 **
## ms_dx_condensedMS, Subtype Not Specified 0.99048
## race_ethnicity_cleanBlack Or African American 0.00785 **
## race_ethnicity_cleanHispanic or Latino 0.39464
## race_ethnicity_cleanWhite Not Hispanic 0.22407
## race_ethnicity_cleanOther/Unknown/Declined 0.16045
## clean_sexMale 0.80938
## clean_sexNon-Binary 0.31723
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1861 on 51 degrees of freedom
## Multiple R-squared: 0.8198, Adjusted R-squared: 0.781
## F-statistic: 21.1 on 11 and 51 DF, p-value: 2.86e-15
# left turns only
home_l_df$sqrt_delta_pix_h_rel_median_pose_hv <- sqrt(home_l_df$delta_pix_h_rel_median_pose_hv)
metric_regression(home_l_df, t25fw_log, sqrt_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 3 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.68453 -0.19988 -0.07443 0.16788 0.86712
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.4121 0.3286 7.340 8.54e-08 ***
## sqrt_delta_pix_h_rel_median_pose_hv -1.4200 0.6021 -2.358 0.0262 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3927 on 26 degrees of freedom
## (3 observations deleted due to missingness)
## Multiple R-squared: 0.1762, Adjusted R-squared: 0.1445
## F-statistic: 5.562 on 1 and 26 DF, p-value: 0.02616
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.53351 -0.07381 0.00000 0.11497 0.49009
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.081223 0.395401 2.734
## sqrt_delta_pix_h_rel_median_pose_hv 0.018925 0.481819 0.039
## demoEHR_Age 0.003316 0.005097 0.651
## demoEHR_DiseaseDuration 0.004821 0.009524 0.506
## ms_dx_condensedProgressive MS 0.808109 0.179979 4.490
## ms_dx_condensedMS, Subtype Not Specified -0.031370 0.229749 -0.137
## race_ethnicity_cleanBlack Or African American 1.391236 0.403058 3.452
## race_ethnicity_cleanHispanic or Latino 0.381638 0.425005 0.898
## race_ethnicity_cleanWhite Not Hispanic 0.276421 0.283458 0.975
## race_ethnicity_cleanOther/Unknown/Declined 0.086125 0.309694 0.278
## clean_sexMale -0.187213 0.167330 -1.119
## clean_sexNon-Binary 0.048441 0.283151 0.171
## Pr(>|t|)
## (Intercept) 0.014695 *
## sqrt_delta_pix_h_rel_median_pose_hv 0.969154
## demoEHR_Age 0.524592
## demoEHR_DiseaseDuration 0.619596
## ms_dx_condensedProgressive MS 0.000371 ***
## ms_dx_condensedMS, Subtype Not Specified 0.893097
## race_ethnicity_cleanBlack Or African American 0.003282 **
## race_ethnicity_cleanHispanic or Latino 0.382520
## race_ethnicity_cleanWhite Not Hispanic 0.343988
## race_ethnicity_cleanOther/Unknown/Declined 0.784499
## clean_sexMale 0.279730
## clean_sexNon-Binary 0.866308
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.258 on 16 degrees of freedom
## (3 observations deleted due to missingness)
## Multiple R-squared: 0.7811, Adjusted R-squared: 0.6306
## F-statistic: 5.19 on 11 and 16 DF, p-value: 0.001624
# right turns only
home_r_df$sqrt_delta_pix_h_rel_median_pose_hv <- sqrt(home_r_df$delta_pix_h_rel_median_pose_hv)
metric_regression(home_r_df, t25fw_log, sqrt_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 1 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.50118 -0.23370 -0.04484 0.13452 0.87680
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3641 0.2707 8.734 1.29e-09 ***
## sqrt_delta_pix_h_rel_median_pose_hv -1.3483 0.4797 -2.811 0.00877 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3155 on 29 degrees of freedom
## (1 observation deleted due to missingness)
## Multiple R-squared: 0.2141, Adjusted R-squared: 0.187
## F-statistic: 7.9 on 1 and 29 DF, p-value: 0.008769
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.45621 -0.10825 -0.00677 0.13491 0.46250
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.492417 0.372937 4.002
## sqrt_delta_pix_h_rel_median_pose_hv -0.352309 0.587835 -0.599
## demoEHR_Age 0.003921 0.004984 0.787
## demoEHR_DiseaseDuration 0.001737 0.008362 0.208
## ms_dx_condensedProgressive MS 0.547264 0.172796 3.167
## ms_dx_condensedMS, Subtype Not Specified 0.045149 0.235023 0.192
## race_ethnicity_cleanHispanic or Latino 0.240353 0.397192 0.605
## race_ethnicity_cleanWhite Not Hispanic 0.114290 0.220884 0.517
## race_ethnicity_cleanOther/Unknown/Declined -0.060355 0.267268 -0.226
## clean_sexMale -0.312761 0.172572 -1.812
## clean_sexNon-Binary -0.016070 0.292315 -0.055
## Pr(>|t|)
## (Intercept) 0.000701 ***
## sqrt_delta_pix_h_rel_median_pose_hv 0.555680
## demoEHR_Age 0.440656
## demoEHR_DiseaseDuration 0.837514
## ms_dx_condensedProgressive MS 0.004846 **
## ms_dx_condensedMS, Subtype Not Specified 0.849597
## race_ethnicity_cleanHispanic or Latino 0.551897
## race_ethnicity_cleanWhite Not Hispanic 0.610539
## race_ethnicity_cleanOther/Unknown/Declined 0.823631
## clean_sexMale 0.084973 .
## clean_sexNon-Binary 0.956704
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2696 on 20 degrees of freedom
## (1 observation deleted due to missingness)
## Multiple R-squared: 0.6044, Adjusted R-squared: 0.4066
## F-statistic: 3.056 on 10 and 20 DF, p-value: 0.01605
# unique ids
home_uniqueid_df$sqrt_delta_pix_h_rel_median_pose_hv <- sqrt(home_uniqueid_df$delta_pix_h_rel_median_pose_hv)
metric_regression(home_uniqueid_df, t25fw_log, sqrt_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_uniqueid_df"## Warning: Removed 3 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.65663 -0.22264 -0.08274 0.17771 0.88134
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3751 0.3422 6.940 2.83e-07 ***
## sqrt_delta_pix_h_rel_median_pose_hv -1.3678 0.6446 -2.122 0.0439 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.403 on 25 degrees of freedom
## (3 observations deleted due to missingness)
## Multiple R-squared: 0.1526, Adjusted R-squared: 0.1187
## F-statistic: 4.503 on 1 and 25 DF, p-value: 0.04393
## [1] "t25fw_log ~ sqrt_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.43106 -0.16370 0.00000 0.08888 0.43942
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.1925813 0.4720607 2.526
## sqrt_delta_pix_h_rel_median_pose_hv -0.0104783 0.5998545 -0.017
## demoEHR_Age 0.0009337 0.0063262 0.148
## demoEHR_DiseaseDuration 0.0013203 0.0111778 0.118
## ms_dx_condensedProgressive MS 0.6145986 0.1874931 3.278
## ms_dx_condensedMS, Subtype Not Specified 0.0869331 0.2715630 0.320
## race_ethnicity_cleanBlack Or African American 1.4833659 0.4822130 3.076
## race_ethnicity_cleanHispanic or Latino 0.5401662 0.5076093 1.064
## race_ethnicity_cleanWhite Not Hispanic 0.3902701 0.3404247 1.146
## race_ethnicity_cleanOther/Unknown/Declined 0.1543245 0.3601969 0.428
## clean_sexMale -0.3916064 0.1857237 -2.109
## clean_sexNon-Binary -0.0516429 0.3329611 -0.155
## Pr(>|t|)
## (Intercept) 0.02326 *
## sqrt_delta_pix_h_rel_median_pose_hv 0.98629
## demoEHR_Age 0.88463
## demoEHR_DiseaseDuration 0.90754
## ms_dx_condensedProgressive MS 0.00508 **
## ms_dx_condensedMS, Subtype Not Specified 0.75329
## race_ethnicity_cleanBlack Or African American 0.00768 **
## race_ethnicity_cleanHispanic or Latino 0.30410
## race_ethnicity_cleanWhite Not Hispanic 0.26958
## race_ethnicity_cleanOther/Unknown/Declined 0.67442
## clean_sexMale 0.05221 .
## clean_sexNon-Binary 0.87881
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3051 on 15 degrees of freedom
## (3 observations deleted due to missingness)
## Multiple R-squared: 0.7086, Adjusted R-squared: 0.4948
## F-statistic: 3.315 on 11 and 15 DF, p-value: 0.01673
Interesting - Vel pixel proxy seems not as good (in this version) in home vs in-person video. Maybe due to variability in recoridng set up at home? In this dataset, demographics seems to explain a lot of variance (maybe small sample size + outliers, even after log transform?)
To do: try again with eye to ankle distance, seemed a lot better in previous versions. Not sure why it’s gotten worse
Univariate - home, stride time –> T25fw
sum(is.finite(home_df$stride_time_median_sec_pose_hv))## [1] 50# Home Videos
metric_regression(home_df, t25fw_log, stride_time_median_sec_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 13 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.45339 -0.22717 0.00424 0.15417 0.53730
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.3659 0.1853 1.974 0.0541 .
## stride_time_median_sec_pose_hv 1.0717 0.1530 7.003 7.28e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2683 on 48 degrees of freedom
## (13 observations deleted due to missingness)
## Multiple R-squared: 0.5053, Adjusted R-squared: 0.495
## F-statistic: 49.04 on 1 and 48 DF, p-value: 7.283e-09
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.37955 -0.12947 -0.02151 0.14532 0.41628
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.396759 0.332484 1.193
## stride_time_median_sec_pose_hv 0.595157 0.243878 2.440
## demoEHR_Age 0.004271 0.003243 1.317
## demoEHR_DiseaseDuration 0.006188 0.005044 1.227
## ms_dx_condensedProgressive MS 0.374928 0.191185 1.961
## ms_dx_condensedMS, Subtype Not Specified -0.123825 0.133685 -0.926
## race_ethnicity_cleanHispanic or Latino 0.331041 0.236839 1.398
## race_ethnicity_cleanWhite Not Hispanic 0.285537 0.162036 1.762
## race_ethnicity_cleanOther/Unknown/Declined 0.069780 0.167938 0.416
## clean_sexMale -0.200774 0.103586 -1.938
## Pr(>|t|)
## (Intercept) 0.2398
## stride_time_median_sec_pose_hv 0.0192 *
## demoEHR_Age 0.1953
## demoEHR_DiseaseDuration 0.2271
## ms_dx_condensedProgressive MS 0.0569 .
## ms_dx_condensedMS, Subtype Not Specified 0.3599
## race_ethnicity_cleanHispanic or Latino 0.1699
## race_ethnicity_cleanWhite Not Hispanic 0.0857 .
## race_ethnicity_cleanOther/Unknown/Declined 0.6800
## clean_sexMale 0.0597 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2052 on 40 degrees of freedom
## (13 observations deleted due to missingness)
## Multiple R-squared: 0.7589, Adjusted R-squared: 0.7047
## F-statistic: 13.99 on 9 and 40 DF, p-value: 7.852e-10
# PWS from mat
metric_regression(home_df, t25fw_log, PWS_stridetimesecmean)## [1] "Data Frame: home_df"
## [1] "t25fw_log ~ PWS_stridetimesecmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47935 -0.22220 -0.03848 0.19794 0.61522
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.4821 0.1288 3.745 0.000403 ***
## PWS_stridetimesecmean 0.9694 0.1029 9.419 1.67e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2559 on 61 degrees of freedom
## Multiple R-squared: 0.5926, Adjusted R-squared: 0.5859
## F-statistic: 88.71 on 1 and 61 DF, p-value: 1.673e-13
## [1] "t25fw_log ~ PWS_stridetimesecmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.41811 -0.10640 0.00000 0.09373 0.38535
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.240538 0.231232 1.040
## PWS_stridetimesecmean 0.813542 0.140701 5.782
## demoEHR_Age 0.008951 0.002509 3.568
## demoEHR_DiseaseDuration 0.004954 0.004142 1.196
## ms_dx_condensedProgressive MS 0.042125 0.133611 0.315
## ms_dx_condensedMS, Subtype Not Specified -0.126098 0.113740 -1.109
## race_ethnicity_cleanBlack Or African American 0.574502 0.203894 2.818
## race_ethnicity_cleanHispanic or Latino -0.047537 0.188192 -0.253
## race_ethnicity_cleanWhite Not Hispanic -0.070969 0.104978 -0.676
## race_ethnicity_cleanOther/Unknown/Declined -0.177090 0.121111 -1.462
## clean_sexMale -0.127546 0.079501 -1.604
## clean_sexNon-Binary 0.099729 0.140663 0.709
## Pr(>|t|)
## (Intercept) 0.303134
## PWS_stridetimesecmean 4.49e-07 ***
## demoEHR_Age 0.000793 ***
## demoEHR_DiseaseDuration 0.237224
## ms_dx_condensedProgressive MS 0.753833
## ms_dx_condensedMS, Subtype Not Specified 0.272784
## race_ethnicity_cleanBlack Or African American 0.006866 **
## race_ethnicity_cleanHispanic or Latino 0.801592
## race_ethnicity_cleanWhite Not Hispanic 0.502070
## race_ethnicity_cleanOther/Unknown/Declined 0.149818
## clean_sexMale 0.114817
## clean_sexNon-Binary 0.481558
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.182 on 51 degrees of freedom
## Multiple R-squared: 0.8278, Adjusted R-squared: 0.7906
## F-statistic: 22.28 on 11 and 51 DF, p-value: 9.46e-16
# left only - video
metric_regression(home_l_df, t25fw_log, stride_time_median_sec_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 6 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44480 -0.19687 0.01752 0.13429 0.47149
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.3098 0.2449 1.265 0.219
## stride_time_median_sec_pose_hv 1.1124 0.2013 5.527 1.28e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2568 on 23 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.5704, Adjusted R-squared: 0.5518
## F-statistic: 30.54 on 1 and 23 DF, p-value: 1.275e-05
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.22598 -0.11669 -0.05208 0.16436 0.37668
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) -0.152921 0.585934 -0.261
## stride_time_median_sec_pose_hv 1.032963 0.431706 2.393
## demoEHR_Age 0.005565 0.004749 1.172
## demoEHR_DiseaseDuration 0.006353 0.007777 0.817
## ms_dx_condensedProgressive MS 0.023369 0.348729 0.067
## ms_dx_condensedMS, Subtype Not Specified -0.219404 0.204168 -1.075
## race_ethnicity_cleanHispanic or Latino 0.297629 0.355074 0.838
## race_ethnicity_cleanWhite Not Hispanic 0.292543 0.240827 1.215
## race_ethnicity_cleanOther/Unknown/Declined 0.092405 0.257604 0.359
## clean_sexMale -0.233820 0.140859 -1.660
## Pr(>|t|)
## (Intercept) 0.7977
## stride_time_median_sec_pose_hv 0.0303 *
## demoEHR_Age 0.2596
## demoEHR_DiseaseDuration 0.4268
## ms_dx_condensedProgressive MS 0.9475
## ms_dx_condensedMS, Subtype Not Specified 0.2995
## race_ethnicity_cleanHispanic or Latino 0.4151
## race_ethnicity_cleanWhite Not Hispanic 0.2432
## race_ethnicity_cleanOther/Unknown/Declined 0.7248
## clean_sexMale 0.1177
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2162 on 15 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.8013, Adjusted R-squared: 0.6822
## F-statistic: 6.723 on 9 and 15 DF, p-value: 0.0006798
# right only - video
metric_regression(home_r_df, t25fw_log, stride_time_median_sec_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 7 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.46165 -0.23398 -0.00483 0.14562 0.54980
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.4262 0.2903 1.468 0.155714
## stride_time_median_sec_pose_hv 1.0271 0.2410 4.262 0.000293 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2897 on 23 degrees of freedom
## (7 observations deleted due to missingness)
## Multiple R-squared: 0.4412, Adjusted R-squared: 0.4169
## F-statistic: 18.16 on 1 and 23 DF, p-value: 0.0002932
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.4354 -0.1159 -0.0196 0.1278 0.4333
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.649731 0.512776 1.267
## stride_time_median_sec_pose_hv 0.368614 0.379722 0.971
## demoEHR_Age 0.004578 0.005865 0.781
## demoEHR_DiseaseDuration 0.004793 0.008501 0.564
## ms_dx_condensedProgressive MS 0.542075 0.285735 1.897
## ms_dx_condensedMS, Subtype Not Specified -0.102353 0.231086 -0.443
## race_ethnicity_cleanHispanic or Latino 0.344817 0.398923 0.864
## race_ethnicity_cleanWhite Not Hispanic 0.275847 0.274976 1.003
## race_ethnicity_cleanOther/Unknown/Declined 0.077020 0.279189 0.276
## clean_sexMale -0.154087 0.200237 -0.770
## Pr(>|t|)
## (Intercept) 0.2244
## stride_time_median_sec_pose_hv 0.3471
## demoEHR_Age 0.4472
## demoEHR_DiseaseDuration 0.5812
## ms_dx_condensedProgressive MS 0.0772 .
## ms_dx_condensedMS, Subtype Not Specified 0.6641
## race_ethnicity_cleanHispanic or Latino 0.4010
## race_ethnicity_cleanWhite Not Hispanic 0.3317
## race_ethnicity_cleanOther/Unknown/Declined 0.7864
## clean_sexMale 0.4535
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2449 on 15 degrees of freedom
## (7 observations deleted due to missingness)
## Multiple R-squared: 0.7395, Adjusted R-squared: 0.5833
## F-statistic: 4.732 on 9 and 15 DF, p-value: 0.004091
# unique IDs - video
metric_regression(home_uniqueid_df, t25fw_log, stride_time_median_sec_pose_hv)## [1] "Data Frame: home_uniqueid_df"## Warning: Removed 8 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44506 -0.21030 0.01718 0.17693 0.47123
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.2927 0.2651 1.104 0.283
## stride_time_median_sec_pose_hv 1.1272 0.2147 5.250 3.88e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2681 on 20 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.5795, Adjusted R-squared: 0.5585
## F-statistic: 27.57 on 1 and 20 DF, p-value: 3.876e-05
## [1] "t25fw_log ~ stride_time_median_sec_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.23616 -0.10344 -0.06526 0.14159 0.34328
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) -0.470725 0.671174 -0.701
## stride_time_median_sec_pose_hv 1.272269 0.486587 2.615
## demoEHR_Age 0.007124 0.005516 1.291
## demoEHR_DiseaseDuration 0.005249 0.008169 0.643
## ms_dx_condensedProgressive MS -0.169488 0.388935 -0.436
## ms_dx_condensedMS, Subtype Not Specified -0.273667 0.223933 -1.222
## race_ethnicity_cleanHispanic or Latino 0.274901 0.381617 0.720
## race_ethnicity_cleanWhite Not Hispanic 0.290239 0.260403 1.115
## race_ethnicity_cleanOther/Unknown/Declined 0.108746 0.266573 0.408
## clean_sexMale -0.240160 0.155160 -1.548
## Pr(>|t|)
## (Intercept) 0.4965
## stride_time_median_sec_pose_hv 0.0226 *
## demoEHR_Age 0.2209
## demoEHR_DiseaseDuration 0.5326
## ms_dx_condensedProgressive MS 0.6707
## ms_dx_condensedMS, Subtype Not Specified 0.2451
## race_ethnicity_cleanHispanic or Latino 0.4851
## race_ethnicity_cleanWhite Not Hispanic 0.2869
## race_ethnicity_cleanOther/Unknown/Declined 0.6905
## clean_sexMale 0.1476
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2258 on 12 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.8212, Adjusted R-squared: 0.6871
## F-statistic: 6.123 on 9 and 12 DF, p-value: 0.002479
Univariate - home, cadence –> T25fw
# Home Videos
sum(is.finite(home_df$mean_cadence_step_per_min_pose_hv))## [1] 51# cadence model
metric_regression(home_df, t25fw_log, mean_cadence_step_per_min_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 12 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.62314 -0.29204 -0.02755 0.21058 0.78453
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.655426 0.264010 10.058 1.66e-13 ***
## mean_cadence_step_per_min_pose_hv -0.010018 0.002537 -3.949 0.000251 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3317 on 49 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.2414, Adjusted R-squared: 0.226
## F-statistic: 15.6 on 1 and 49 DF, p-value: 0.0002509
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51563 -0.09908 -0.02010 0.14415 0.42519
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.425855 0.287774 4.955
## mean_cadence_step_per_min_pose_hv -0.002999 0.002054 -1.460
## demoEHR_Age 0.002676 0.003197 0.837
## demoEHR_DiseaseDuration 0.006096 0.005122 1.190
## ms_dx_condensedProgressive MS 0.676573 0.122683 5.515
## ms_dx_condensedMS, Subtype Not Specified -0.059394 0.133070 -0.446
## race_ethnicity_cleanHispanic or Latino 0.368590 0.244533 1.507
## race_ethnicity_cleanWhite Not Hispanic 0.317195 0.167575 1.893
## race_ethnicity_cleanOther/Unknown/Declined 0.078784 0.173942 0.453
## clean_sexMale -0.227672 0.098397 -2.314
## Pr(>|t|)
## (Intercept) 1.30e-05 ***
## mean_cadence_step_per_min_pose_hv 0.1518
## demoEHR_Age 0.4074
## demoEHR_DiseaseDuration 0.2409
## ms_dx_condensedProgressive MS 2.11e-06 ***
## ms_dx_condensedMS, Subtype Not Specified 0.6577
## race_ethnicity_cleanHispanic or Latino 0.1394
## race_ethnicity_cleanWhite Not Hispanic 0.0655 .
## race_ethnicity_cleanOther/Unknown/Declined 0.6530
## clean_sexMale 0.0258 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2126 on 41 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.7394, Adjusted R-squared: 0.6822
## F-statistic: 12.92 on 9 and 41 DF, p-value: 1.893e-09
# ground truth cadence from mat
metric_regression(home_df, t25fw_log, PWS_cadencestepsminmean)## [1] "Data Frame: home_df"
## [1] "t25fw_log ~ PWS_cadencestepsminmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.54113 -0.22498 -0.00353 0.20685 0.57072
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.273424 0.205824 15.904 < 2e-16 ***
## PWS_cadencestepsminmean -0.015691 0.001967 -7.976 4.84e-11 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2805 on 61 degrees of freedom
## Multiple R-squared: 0.5105, Adjusted R-squared: 0.5025
## F-statistic: 63.61 on 1 and 61 DF, p-value: 4.842e-11
## [1] "t25fw_log ~ PWS_cadencestepsminmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.4340 -0.1191 0.0000 0.1292 0.3680
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.265792 0.294904 7.683
## PWS_cadencestepsminmean -0.009766 0.002617 -3.731
## demoEHR_Age 0.007335 0.002829 2.593
## demoEHR_DiseaseDuration 0.006899 0.004896 1.409
## ms_dx_condensedProgressive MS 0.288894 0.135298 2.135
## ms_dx_condensedMS, Subtype Not Specified -0.097336 0.130032 -0.749
## race_ethnicity_cleanBlack Or African American 0.652933 0.244291 2.673
## race_ethnicity_cleanHispanic or Latino -0.063416 0.219264 -0.289
## race_ethnicity_cleanWhite Not Hispanic -0.064278 0.121292 -0.530
## race_ethnicity_cleanOther/Unknown/Declined -0.197045 0.138981 -1.418
## clean_sexMale -0.171942 0.090579 -1.898
## clean_sexNon-Binary 0.099414 0.161590 0.615
## Pr(>|t|)
## (Intercept) 4.56e-10 ***
## PWS_cadencestepsminmean 0.00048 ***
## demoEHR_Age 0.01239 *
## demoEHR_DiseaseDuration 0.16485
## ms_dx_condensedProgressive MS 0.03757 *
## ms_dx_condensedMS, Subtype Not Specified 0.45757
## race_ethnicity_cleanBlack Or African American 0.01008 *
## race_ethnicity_cleanHispanic or Latino 0.77358
## race_ethnicity_cleanWhite Not Hispanic 0.59845
## race_ethnicity_cleanOther/Unknown/Declined 0.16233
## clean_sexMale 0.06333 .
## clean_sexNon-Binary 0.54114
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2075 on 51 degrees of freedom
## Multiple R-squared: 0.776, Adjusted R-squared: 0.7277
## F-statistic: 16.06 on 11 and 51 DF, p-value: 5.832e-13
# Left turns only
metric_regression(home_l_df, t25fw_log, mean_cadence_step_per_min_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 6 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51302 -0.25205 -0.01996 0.25132 0.71374
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.86342 0.36578 7.828 6.21e-08 ***
## mean_cadence_step_per_min_pose_hv -0.01178 0.00345 -3.415 0.00237 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3191 on 23 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.3365, Adjusted R-squared: 0.3076
## F-statistic: 11.66 on 1 and 23 DF, p-value: 0.00237
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47903 -0.14227 -0.00492 0.16519 0.38651
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.607878 0.478264 3.362
## mean_cadence_step_per_min_pose_hv -0.004678 0.003513 -1.332
## demoEHR_Age 0.003291 0.005137 0.641
## demoEHR_DiseaseDuration 0.005234 0.008678 0.603
## ms_dx_condensedProgressive MS 0.599228 0.211044 2.839
## ms_dx_condensedMS, Subtype Not Specified -0.097642 0.215824 -0.452
## race_ethnicity_cleanHispanic or Latino 0.300032 0.396547 0.757
## race_ethnicity_cleanWhite Not Hispanic 0.309813 0.267730 1.157
## race_ethnicity_cleanOther/Unknown/Declined 0.046571 0.286849 0.162
## clean_sexMale -0.219844 0.156362 -1.406
## Pr(>|t|)
## (Intercept) 0.00428 **
## mean_cadence_step_per_min_pose_hv 0.20287
## demoEHR_Age 0.53144
## demoEHR_DiseaseDuration 0.55545
## ms_dx_condensedProgressive MS 0.01243 *
## ms_dx_condensedMS, Subtype Not Specified 0.65744
## race_ethnicity_cleanHispanic or Latino 0.46100
## race_ethnicity_cleanWhite Not Hispanic 0.26529
## race_ethnicity_cleanOther/Unknown/Declined 0.87319
## clean_sexMale 0.18010
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2403 on 15 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.7545, Adjusted R-squared: 0.6073
## F-statistic: 5.123 on 9 and 15 DF, p-value: 0.002781
# right turns only
metric_regression(home_r_df, t25fw_log, mean_cadence_step_per_min_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 6 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.56097 -0.31590 0.00281 0.21244 0.79006
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.478456 0.393513 6.298 1.64e-06 ***
## mean_cadence_step_per_min_pose_hv -0.008487 0.003852 -2.203 0.0374 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3521 on 24 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.1682, Adjusted R-squared: 0.1336
## F-statistic: 4.854 on 1 and 24 DF, p-value: 0.03742
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51783 -0.09212 -0.00925 0.12399 0.42613
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.308072 0.477983 2.737
## mean_cadence_step_per_min_pose_hv -0.001921 0.003442 -0.558
## demoEHR_Age 0.002537 0.005282 0.480
## demoEHR_DiseaseDuration 0.005978 0.008251 0.725
## ms_dx_condensedProgressive MS 0.721024 0.193542 3.725
## ms_dx_condensedMS, Subtype Not Specified -0.039601 0.215730 -0.184
## race_ethnicity_cleanHispanic or Latino 0.404191 0.397255 1.017
## race_ethnicity_cleanWhite Not Hispanic 0.319485 0.273685 1.167
## race_ethnicity_cleanOther/Unknown/Declined 0.103026 0.277190 0.372
## clean_sexMale -0.228897 0.162461 -1.409
## Pr(>|t|)
## (Intercept) 0.01463 *
## mean_cadence_step_per_min_pose_hv 0.58447
## demoEHR_Age 0.63750
## demoEHR_DiseaseDuration 0.47917
## ms_dx_condensedProgressive MS 0.00184 **
## ms_dx_condensedMS, Subtype Not Specified 0.85666
## race_ethnicity_cleanHispanic or Latino 0.32407
## race_ethnicity_cleanWhite Not Hispanic 0.26017
## race_ethnicity_cleanOther/Unknown/Declined 0.71501
## clean_sexMale 0.17799
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2449 on 16 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.7318, Adjusted R-squared: 0.5809
## F-statistic: 4.851 on 9 and 16 DF, p-value: 0.003028
# unique IDs
metric_regression(home_uniqueid_df, t25fw_log, mean_cadence_step_per_min_pose_hv)## [1] "Data Frame: home_uniqueid_df"## Warning: Removed 8 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.55032 -0.26840 -0.00729 0.23992 0.66189
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.964594 0.378715 7.828 1.63e-07 ***
## mean_cadence_step_per_min_pose_hv -0.012415 0.003522 -3.525 0.00213 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3248 on 20 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.3832, Adjusted R-squared: 0.3524
## F-statistic: 12.43 on 1 and 20 DF, p-value: 0.002128
## [1] "t25fw_log ~ mean_cadence_step_per_min_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.46879 -0.13521 0.00492 0.13954 0.33860
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.730998 0.551000 3.142
## mean_cadence_step_per_min_pose_hv -0.005597 0.004070 -1.375
## demoEHR_Age 0.002959 0.006091 0.486
## demoEHR_DiseaseDuration 0.002639 0.009586 0.275
## ms_dx_condensedProgressive MS 0.529096 0.246963 2.142
## ms_dx_condensedMS, Subtype Not Specified -0.092789 0.241584 -0.384
## race_ethnicity_cleanHispanic or Latino 0.353918 0.442291 0.800
## race_ethnicity_cleanWhite Not Hispanic 0.364105 0.303324 1.200
## race_ethnicity_cleanOther/Unknown/Declined 0.118661 0.310372 0.382
## clean_sexMale -0.260934 0.181813 -1.435
## Pr(>|t|)
## (Intercept) 0.00851 **
## mean_cadence_step_per_min_pose_hv 0.19422
## demoEHR_Age 0.63579
## demoEHR_DiseaseDuration 0.78778
## ms_dx_condensedProgressive MS 0.05336 .
## ms_dx_condensedMS, Subtype Not Specified 0.70763
## race_ethnicity_cleanHispanic or Latino 0.43915
## race_ethnicity_cleanWhite Not Hispanic 0.25315
## race_ethnicity_cleanOther/Unknown/Declined 0.70891
## clean_sexMale 0.17679
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2629 on 12 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.7575, Adjusted R-squared: 0.5756
## F-statistic: 4.165 on 9 and 12 DF, p-value: 0.01225
Univariate - home, stride width –> T25fw
# Home Videos
sum(is.finite(home_df$stride_width_median_cm_pose_hv))## [1] 51## model
metric_regression(home_df, t25fw_log, stride_width_median_cm_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 12 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51421 -0.23980 0.02417 0.16497 0.83045
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.51023 0.22738 2.244 0.0294 *
## stride_width_median_cm_pose_hv 0.08809 0.01757 5.013 7.41e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3097 on 49 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.339, Adjusted R-squared: 0.3255
## F-statistic: 25.13 on 1 and 49 DF, p-value: 7.408e-06
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47343 -0.05906 0.00437 0.09373 0.32221
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.553533 0.206149 2.685
## stride_width_median_cm_pose_hv 0.050129 0.012259 4.089
## demoEHR_Age 0.001591 0.002778 0.573
## demoEHR_DiseaseDuration 0.005727 0.004407 1.299
## ms_dx_condensedProgressive MS 0.639870 0.092362 6.928
## ms_dx_condensedMS, Subtype Not Specified -0.160131 0.118131 -1.356
## race_ethnicity_cleanHispanic or Latino 0.287112 0.212499 1.351
## race_ethnicity_cleanWhite Not Hispanic 0.312424 0.144667 2.160
## race_ethnicity_cleanOther/Unknown/Declined 0.111080 0.150150 0.740
## clean_sexMale -0.217965 0.084440 -2.581
## Pr(>|t|)
## (Intercept) 0.010417 *
## stride_width_median_cm_pose_hv 0.000197 ***
## demoEHR_Age 0.570079
## demoEHR_DiseaseDuration 0.201041
## ms_dx_condensedProgressive MS 2.08e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.182666
## race_ethnicity_cleanHispanic or Latino 0.184065
## race_ethnicity_cleanWhite Not Hispanic 0.036703 *
## race_ethnicity_cleanOther/Unknown/Declined 0.463640
## clean_sexMale 0.013517 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1838 on 41 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.8052, Adjusted R-squared: 0.7625
## F-statistic: 18.84 on 9 and 41 DF, p-value: 6.384e-12
## pressure mat stride width model
metric_regression(home_df, t25fw_log, PWS_stridewidthcmmean)## [1] "Data Frame: home_df"
## [1] "t25fw_log ~ PWS_stridewidthcmmean"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.59489 -0.21569 -0.05045 0.10379 1.04050
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.36895 0.15466 8.851 1.53e-12 ***
## PWS_stridewidthcmmean 0.03345 0.01708 1.958 0.0548 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3889 on 61 degrees of freedom
## Multiple R-squared: 0.05914, Adjusted R-squared: 0.04372
## F-statistic: 3.834 on 1 and 61 DF, p-value: 0.05479
## [1] "t25fw_log ~ PWS_stridewidthcmmean + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.57079 -0.09658 0.00000 0.11307 0.41540
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.188671 0.199078 5.971
## PWS_stridewidthcmmean 0.016321 0.012149 1.343
## demoEHR_Age 0.003769 0.003033 1.242
## demoEHR_DiseaseDuration 0.003020 0.005312 0.569
## ms_dx_condensedProgressive MS 0.681621 0.099350 6.861
## ms_dx_condensedMS, Subtype Not Specified 0.056866 0.146133 0.389
## race_ethnicity_cleanBlack Or African American 1.086501 0.231640 4.690
## race_ethnicity_cleanHispanic or Latino 0.123461 0.235517 0.524
## race_ethnicity_cleanWhite Not Hispanic 0.061066 0.130208 0.469
## race_ethnicity_cleanOther/Unknown/Declined -0.145615 0.153262 -0.950
## clean_sexMale -0.278153 0.094439 -2.945
## clean_sexNon-Binary 0.025798 0.178107 0.145
## Pr(>|t|)
## (Intercept) 2.28e-07 ***
## PWS_stridewidthcmmean 0.18510
## demoEHR_Age 0.21979
## demoEHR_DiseaseDuration 0.57217
## ms_dx_condensedProgressive MS 9.03e-09 ***
## ms_dx_condensedMS, Subtype Not Specified 0.69879
## race_ethnicity_cleanBlack Or African American 2.08e-05 ***
## race_ethnicity_cleanHispanic or Latino 0.60240
## race_ethnicity_cleanWhite Not Hispanic 0.64108
## race_ethnicity_cleanOther/Unknown/Declined 0.34654
## clean_sexMale 0.00485 **
## clean_sexNon-Binary 0.88540
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2301 on 51 degrees of freedom
## Multiple R-squared: 0.7246, Adjusted R-squared: 0.6652
## F-statistic: 12.2 on 11 and 51 DF, p-value: 8.445e-11
# left only
metric_regression(home_l_df, t25fw_log, stride_width_median_cm_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 6 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.43106 -0.27372 0.00517 0.17680 0.83812
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.73046 0.39970 1.828 0.0806 .
## stride_width_median_cm_pose_hv 0.07175 0.03126 2.295 0.0312 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3534 on 23 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.1864, Adjusted R-squared: 0.151
## F-statistic: 5.269 on 1 and 23 DF, p-value: 0.03117
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47673 -0.04122 0.00401 0.07947 0.26128
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.4898646 0.3035972 1.614
## stride_width_median_cm_pose_hv 0.0633025 0.0197018 3.213
## demoEHR_Age -0.0002653 0.0042833 -0.062
## demoEHR_DiseaseDuration 0.0099258 0.0071279 1.393
## ms_dx_condensedProgressive MS 0.6918538 0.1321152 5.237
## ms_dx_condensedMS, Subtype Not Specified -0.1501074 0.1759680 -0.853
## race_ethnicity_cleanHispanic or Latino 0.2319664 0.3228908 0.718
## race_ethnicity_cleanWhite Not Hispanic 0.2880571 0.2178812 1.322
## race_ethnicity_cleanOther/Unknown/Declined -0.0093048 0.2343029 -0.040
## clean_sexMale -0.3245555 0.1320527 -2.458
## Pr(>|t|)
## (Intercept) 0.12746
## stride_width_median_cm_pose_hv 0.00581 **
## demoEHR_Age 0.95143
## demoEHR_DiseaseDuration 0.18406
## ms_dx_condensedProgressive MS 0.00010 ***
## ms_dx_condensedMS, Subtype Not Specified 0.40705
## race_ethnicity_cleanHispanic or Latino 0.48355
## race_ethnicity_cleanWhite Not Hispanic 0.20595
## race_ethnicity_cleanOther/Unknown/Declined 0.96885
## clean_sexMale 0.02663 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1956 on 15 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.8374, Adjusted R-squared: 0.7399
## F-statistic: 8.584 on 9 and 15 DF, p-value: 0.0001725
# right only
metric_regression(home_r_df, t25fw_log, stride_width_median_cm_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 6 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.50540 -0.18496 0.03658 0.12051 0.81958
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.34341 0.26231 1.309 0.203
## stride_width_median_cm_pose_hv 0.10001 0.02005 4.988 4.28e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2705 on 24 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.509, Adjusted R-squared: 0.4886
## F-statistic: 24.88 on 1 and 24 DF, p-value: 4.28e-05
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.39743 -0.03506 0.01639 0.04885 0.37607
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.425543 0.363698 1.170
## stride_width_median_cm_pose_hv 0.055640 0.021719 2.562
## demoEHR_Age 0.002550 0.004407 0.579
## demoEHR_DiseaseDuration 0.003172 0.006795 0.467
## ms_dx_condensedProgressive MS 0.548278 0.164310 3.337
## ms_dx_condensedMS, Subtype Not Specified -0.214161 0.195483 -1.096
## race_ethnicity_cleanHispanic or Latino 0.287874 0.339396 0.848
## race_ethnicity_cleanWhite Not Hispanic 0.334120 0.232009 1.440
## race_ethnicity_cleanOther/Unknown/Declined 0.198416 0.238176 0.833
## clean_sexMale -0.126195 0.139213 -0.906
## Pr(>|t|)
## (Intercept) 0.25912
## stride_width_median_cm_pose_hv 0.02090 *
## demoEHR_Age 0.57089
## demoEHR_DiseaseDuration 0.64686
## ms_dx_condensedProgressive MS 0.00418 **
## ms_dx_condensedMS, Subtype Not Specified 0.28949
## race_ethnicity_cleanHispanic or Latino 0.40884
## race_ethnicity_cleanWhite Not Hispanic 0.16911
## race_ethnicity_cleanOther/Unknown/Declined 0.41707
## clean_sexMale 0.37813
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2082 on 16 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.8061, Adjusted R-squared: 0.697
## F-statistic: 7.391 on 9 and 16 DF, p-value: 0.0003013
# unique IDs
metric_regression(home_uniqueid_df, t25fw_log, stride_width_median_cm_pose_hv)## [1] "Data Frame: home_uniqueid_df"## Warning: Removed 8 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.42575 -0.28031 0.01273 0.19963 0.81098
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.60218 0.45328 1.328 0.1990
## stride_width_median_cm_pose_hv 0.08196 0.03486 2.351 0.0291 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.366 on 20 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.2166, Adjusted R-squared: 0.1774
## F-statistic: 5.529 on 1 and 20 DF, p-value: 0.02906
## [1] "t25fw_log ~ stride_width_median_cm_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.41032 -0.08328 -0.00005 0.14000 0.25147
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.422694 0.359137 1.177
## stride_width_median_cm_pose_hv 0.067558 0.022977 2.940
## demoEHR_Age 0.001325 0.005003 0.265
## demoEHR_DiseaseDuration 0.001380 0.007862 0.176
## ms_dx_condensedProgressive MS 0.675926 0.147378 4.586
## ms_dx_condensedMS, Subtype Not Specified -0.102221 0.196263 -0.521
## race_ethnicity_cleanHispanic or Latino 0.355176 0.361294 0.983
## race_ethnicity_cleanWhite Not Hispanic 0.291171 0.248645 1.171
## race_ethnicity_cleanOther/Unknown/Declined 0.109104 0.254620 0.428
## clean_sexMale -0.324093 0.151407 -2.141
## Pr(>|t|)
## (Intercept) 0.262025
## stride_width_median_cm_pose_hv 0.012366 *
## demoEHR_Age 0.795602
## demoEHR_DiseaseDuration 0.863570
## ms_dx_condensedProgressive MS 0.000626 ***
## ms_dx_condensedMS, Subtype Not Specified 0.611957
## race_ethnicity_cleanHispanic or Latino 0.344981
## race_ethnicity_cleanWhite Not Hispanic 0.264314
## race_ethnicity_cleanOther/Unknown/Declined 0.675879
## clean_sexMale 0.053540 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2156 on 12 degrees of freedom
## (8 observations deleted due to missingness)
## Multiple R-squared: 0.8368, Adjusted R-squared: 0.7145
## F-statistic: 6.839 on 9 and 12 DF, p-value: 0.00151
Interesting Stride width associated with T25fw in home videos, but not in person…?
Univariate - home, Stance Time –> T25fw
Stance/swing/double/single support measures not calculated for all participants
# Home Video
sum(is.finite(home_df$foot1_stance_time_mean_pose_hv))## [1] 31#video model
metric_regression(home_df, t25fw_log, foot1_stance_time_mean_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 32 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.41369 -0.21596 -0.02146 0.22394 0.50692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.8684 0.1333 6.516 3.91e-07 ***
## foot1_stance_time_mean_pose_hv 0.9180 0.1425 6.441 4.79e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2601 on 29 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.5886, Adjusted R-squared: 0.5744
## F-statistic: 41.48 on 1 and 29 DF, p-value: 4.789e-07
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.39587 -0.14503 -0.02896 0.13481 0.45123
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.916236 0.328072 2.793
## foot1_stance_time_mean_pose_hv 0.582100 0.245299 2.373
## demoEHR_Age 0.003896 0.004971 0.784
## demoEHR_DiseaseDuration 0.006127 0.008551 0.716
## ms_dx_condensedProgressive MS 0.281338 0.233640 1.204
## ms_dx_condensedMS, Subtype Not Specified -0.133046 0.165995 -0.802
## race_ethnicity_cleanOther/Unknown/Declined -0.204713 0.165585 -1.236
## clean_sexMale -0.195809 0.134007 -1.461
## Pr(>|t|)
## (Intercept) 0.0103 *
## foot1_stance_time_mean_pose_hv 0.0264 *
## demoEHR_Age 0.4412
## demoEHR_DiseaseDuration 0.4809
## ms_dx_condensedProgressive MS 0.2408
## ms_dx_condensedMS, Subtype Not Specified 0.4310
## race_ethnicity_cleanOther/Unknown/Declined 0.2288
## clean_sexMale 0.1575
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2325 on 23 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.7393, Adjusted R-squared: 0.66
## F-statistic: 9.318 on 7 and 23 DF, p-value: 1.866e-05
# no pressure mat value
# left only
metric_regression(home_l_df, t25fw_log, foot1_stance_time_mean_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 18 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.43050 -0.25282 0.09124 0.15191 0.51482
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.7179 0.2750 2.611 0.02423 *
## foot1_stance_time_mean_pose_hv 1.1039 0.3145 3.510 0.00489 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2856 on 11 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.5282, Adjusted R-squared: 0.4854
## F-statistic: 12.32 on 1 and 11 DF, p-value: 0.004887
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## 1 9 13 15 17 22 28
## -8.747e-02 -1.468e-01 1.164e-01 1.601e-01 8.747e-02 4.857e-17 -8.675e-02
## 38 42 52 56 60 62
## 1.468e-01 -1.316e-01 -9.460e-02 3.347e-01 -2.982e-01 6.939e-18
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.888313 0.493821 1.799
## foot1_stance_time_mean_pose_hv 1.186527 0.643712 1.843
## demoEHR_Age 0.001566 0.008696 0.180
## demoEHR_DiseaseDuration -0.015355 0.020965 -0.732
## ms_dx_condensedProgressive MS -0.191498 0.460179 -0.416
## ms_dx_condensedMS, Subtype Not Specified -0.395736 0.318957 -1.241
## race_ethnicity_cleanOther/Unknown/Declined -0.387180 0.301270 -1.285
## clean_sexMale -0.615623 0.260421 -2.364
## Pr(>|t|)
## (Intercept) 0.1320
## foot1_stance_time_mean_pose_hv 0.1246
## demoEHR_Age 0.8642
## demoEHR_DiseaseDuration 0.4968
## ms_dx_condensedProgressive MS 0.6946
## ms_dx_condensedMS, Subtype Not Specified 0.2697
## race_ethnicity_cleanOther/Unknown/Declined 0.2550
## clean_sexMale 0.0644 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2578 on 5 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.8252, Adjusted R-squared: 0.5806
## F-statistic: 3.373 on 7 and 5 DF, p-value: 0.09973
# right only
metric_regression(home_r_df, t25fw_log, foot1_stance_time_mean_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 14 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.36323 -0.20606 -0.01964 0.22125 0.38835
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.9190 0.1573 5.842 2.50e-05 ***
## foot1_stance_time_mean_pose_hv 0.8579 0.1611 5.325 6.84e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2534 on 16 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.6393, Adjusted R-squared: 0.6167
## F-statistic: 28.36 on 1 and 16 DF, p-value: 6.837e-05
## [1] "t25fw_log ~ foot1_stance_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.41999 -0.13309 0.01339 0.13324 0.33547
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.310984 0.630472 0.493
## foot1_stance_time_mean_pose_hv 0.804728 0.376455 2.138
## demoEHR_Age 0.013582 0.009516 1.427
## demoEHR_DiseaseDuration -0.002743 0.013076 -0.210
## ms_dx_condensedProgressive MS 0.053945 0.394634 0.137
## ms_dx_condensedMS, Subtype Not Specified -0.240588 0.256620 -0.938
## race_ethnicity_cleanOther/Unknown/Declined -0.005601 0.238902 -0.023
## clean_sexMale 0.079479 0.241806 0.329
## Pr(>|t|)
## (Intercept) 0.6325
## foot1_stance_time_mean_pose_hv 0.0583 .
## demoEHR_Age 0.1840
## demoEHR_DiseaseDuration 0.8381
## ms_dx_condensedProgressive MS 0.8940
## ms_dx_condensedMS, Subtype Not Specified 0.3706
## race_ethnicity_cleanOther/Unknown/Declined 0.9818
## clean_sexMale 0.7492
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2409 on 10 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7963, Adjusted R-squared: 0.6537
## F-statistic: 5.584 on 7 and 10 DF, p-value: 0.007773
# unique IDs - insufficent data? - less than one factor group
# metric_regression(home_uniqueid_df, t25fw_log, foot1_stance_time_mean_pose_hv)Univariate - home, Swing Time –> T25fw
# Home
sum(is.finite(home_df$foot1_swing_time_mean_pose_hv))## [1] 31ggplot(data = home_df, aes(x = foot1_swing_time_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = foot1_swing_per_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
# home all
metric_regression(home_df, t25fw_log, foot1_swing_per_mean_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 32 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.76050 -0.21748 -0.07662 0.21468 0.59516
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.64720 0.26485 9.995 6.68e-11 ***
## foot1_swing_per_mean_pose_hv -0.03008 0.00796 -3.778 0.000728 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.332 on 29 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.3298, Adjusted R-squared: 0.3067
## F-statistic: 14.27 on 1 and 29 DF, p-value: 0.0007284
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44780 -0.12159 -0.05335 0.11970 0.48762
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.668317 0.352202 4.737
## foot1_swing_per_mean_pose_hv -0.006034 0.008650 -0.698
## demoEHR_Age 0.001255 0.005372 0.234
## demoEHR_DiseaseDuration 0.009338 0.009352 0.999
## ms_dx_condensedProgressive MS 0.657984 0.189904 3.465
## ms_dx_condensedMS, Subtype Not Specified -0.093124 0.182417 -0.511
## race_ethnicity_cleanOther/Unknown/Declined -0.197968 0.183884 -1.077
## clean_sexMale -0.239833 0.146364 -1.639
## Pr(>|t|)
## (Intercept) 8.97e-05 ***
## foot1_swing_per_mean_pose_hv 0.4924
## demoEHR_Age 0.8173
## demoEHR_DiseaseDuration 0.3284
## ms_dx_condensedProgressive MS 0.0021 **
## ms_dx_condensedMS, Subtype Not Specified 0.6146
## race_ethnicity_cleanOther/Unknown/Declined 0.2928
## clean_sexMale 0.1149
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2567 on 23 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.6822, Adjusted R-squared: 0.5855
## F-statistic: 7.053 on 7 and 23 DF, p-value: 0.0001523
# left
metric_regression(home_l_df, t25fw_log, foot1_swing_per_mean_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 18 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.43667 -0.14377 -0.04525 0.13613 0.63718
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.91606 0.39550 7.373 1.41e-05 ***
## foot1_swing_per_mean_pose_hv -0.03895 0.01183 -3.292 0.00718 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2951 on 11 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.4963, Adjusted R-squared: 0.4505
## F-statistic: 10.84 on 1 and 11 DF, p-value: 0.007178
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## 1 9 13 15 17 22 28
## -6.902e-02 -3.696e-01 -9.420e-02 5.489e-02 6.902e-02 4.337e-17 -5.772e-02
## 38 42 52 56 60 62
## 3.696e-01 -2.034e-01 -1.271e-02 4.330e-01 -1.199e-01 -1.214e-17
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.938891 1.291443 1.501
## foot1_swing_per_mean_pose_hv -0.011111 0.028263 -0.393
## demoEHR_Age 0.004564 0.011064 0.413
## demoEHR_DiseaseDuration -0.013171 0.027679 -0.476
## ms_dx_condensedProgressive MS 0.370438 0.516868 0.717
## ms_dx_condensedMS, Subtype Not Specified -0.356351 0.412716 -0.863
## race_ethnicity_cleanOther/Unknown/Declined -0.244240 0.376367 -0.649
## clean_sexMale -0.454526 0.312527 -1.454
## Pr(>|t|)
## (Intercept) 0.194
## foot1_swing_per_mean_pose_hv 0.710
## demoEHR_Age 0.697
## demoEHR_DiseaseDuration 0.654
## ms_dx_condensedProgressive MS 0.506
## ms_dx_condensedMS, Subtype Not Specified 0.427
## race_ethnicity_cleanOther/Unknown/Declined 0.545
## clean_sexMale 0.206
##
## Residual standard error: 0.3291 on 5 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.7153, Adjusted R-squared: 0.3167
## F-statistic: 1.795 on 7 and 5 DF, p-value: 0.2691
# right
metric_regression(home_r_df, t25fw_log, foot1_swing_per_mean_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 14 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.71011 -0.26002 -0.06287 0.24080 0.64492
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.49923 0.36398 6.866 3.79e-06 ***
## foot1_swing_per_mean_pose_hv -0.02500 0.01098 -2.278 0.0368 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3666 on 16 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.2448, Adjusted R-squared: 0.1976
## F-statistic: 5.187 on 1 and 16 DF, p-value: 0.03683
## [1] "t25fw_log ~ foot1_swing_per_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.47783 -0.10059 -0.03943 0.12429 0.47731
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.612267 0.451118 3.574
## foot1_swing_per_mean_pose_hv -0.007423 0.012151 -0.611
## demoEHR_Age 0.003041 0.009574 0.318
## demoEHR_DiseaseDuration 0.009093 0.013840 0.657
## ms_dx_condensedProgressive MS 0.704167 0.271657 2.592
## ms_dx_condensedMS, Subtype Not Specified -0.068797 0.286797 -0.240
## race_ethnicity_cleanOther/Unknown/Declined -0.123715 0.275217 -0.450
## clean_sexMale -0.176651 0.244744 -0.722
## Pr(>|t|)
## (Intercept) 0.00506 **
## foot1_swing_per_mean_pose_hv 0.55490
## demoEHR_Age 0.75732
## demoEHR_DiseaseDuration 0.52602
## ms_dx_condensedProgressive MS 0.02685 *
## ms_dx_condensedMS, Subtype Not Specified 0.81527
## race_ethnicity_cleanOther/Unknown/Declined 0.66264
## clean_sexMale 0.48697
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2854 on 10 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7139, Adjusted R-squared: 0.5136
## F-statistic: 3.564 on 7 and 10 DF, p-value: 0.03434
# unique IDs
#metric_regression(home_uniqueid_df, t25fw_log, foot1_swing_per_mean_pose_hv)Univariate - home, Single and Double Support –> T25fw
# home
sum(is.finite(home_df$foot1_double_support_per_mean_pose_zv))## [1] 0ggplot(data = home_df, aes(x = foot1_double_support_per_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = foot1_ini_double_support_time_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = foot1_single_support_per_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = foot1_term_double_support_time_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = foot1_tot_double_support_time_mean_pose_hv, y = t25fw_log)) +
geom_point()## Warning: Removed 32 rows containing missing values (`geom_point()`).
# home all
metric_regression(home_df, t25fw_log, foot1_ini_double_support_time_mean_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 32 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_ini_double_support_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.55005 -0.20830 0.02674 0.17953 0.57793
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.32828 0.08991 14.774 4.97e-15
## foot1_ini_double_support_time_mean_pose_hv 1.25675 0.26164 4.803 4.38e-05
##
## (Intercept) ***
## foot1_ini_double_support_time_mean_pose_hv ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3026 on 29 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.4431, Adjusted R-squared: 0.4239
## F-statistic: 23.07 on 1 and 29 DF, p-value: 4.381e-05
## [1] "t25fw_log ~ foot1_ini_double_support_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.41357 -0.14148 -0.04077 0.13736 0.45177
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.335203 0.265041 5.038
## foot1_ini_double_support_time_mean_pose_hv 0.445953 0.334721 1.332
## demoEHR_Age 0.002312 0.005297 0.436
## demoEHR_DiseaseDuration 0.007079 0.009324 0.759
## ms_dx_condensedProgressive MS 0.544934 0.204024 2.671
## ms_dx_condensedMS, Subtype Not Specified -0.099146 0.177629 -0.558
## race_ethnicity_cleanOther/Unknown/Declined -0.201706 0.178141 -1.132
## clean_sexMale -0.224191 0.143177 -1.566
## Pr(>|t|)
## (Intercept) 4.25e-05 ***
## foot1_ini_double_support_time_mean_pose_hv 0.1958
## demoEHR_Age 0.6666
## demoEHR_DiseaseDuration 0.4554
## ms_dx_condensedProgressive MS 0.0136 *
## ms_dx_condensedMS, Subtype Not Specified 0.5821
## race_ethnicity_cleanOther/Unknown/Declined 0.2692
## clean_sexMale 0.1310
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2499 on 23 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.6987, Adjusted R-squared: 0.607
## F-statistic: 7.62 on 7 and 23 DF, p-value: 8.689e-05
metric_regression(home_df, t25fw_log, foot1_term_double_support_time_mean_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 32 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.71943 -0.22296 -0.00838 0.23779 0.61245
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.37306 0.09222 14.889
## foot1_term_double_support_time_mean_pose_hv 1.40193 0.33723 4.157
## Pr(>|t|)
## (Intercept) 4.06e-15 ***
## foot1_term_double_support_time_mean_pose_hv 0.000261 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.321 on 29 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.3734, Adjusted R-squared: 0.3518
## F-statistic: 17.28 on 1 and 29 DF, p-value: 0.0002607
## [1] "t25fw_log ~ foot1_term_double_support_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44340 -0.13760 -0.03355 0.11151 0.52202
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.325843 0.251176 5.279
## foot1_term_double_support_time_mean_pose_hv 0.596937 0.345816 1.726
## demoEHR_Age 0.002021 0.005073 0.398
## demoEHR_DiseaseDuration 0.009160 0.008806 1.040
## ms_dx_condensedProgressive MS 0.552673 0.174183 3.173
## ms_dx_condensedMS, Subtype Not Specified -0.107572 0.173592 -0.620
## race_ethnicity_cleanOther/Unknown/Declined -0.180918 0.174726 -1.035
## clean_sexMale -0.251581 0.139054 -1.809
## Pr(>|t|)
## (Intercept) 2.34e-05 ***
## foot1_term_double_support_time_mean_pose_hv 0.09773 .
## demoEHR_Age 0.69400
## demoEHR_DiseaseDuration 0.30907
## ms_dx_condensedProgressive MS 0.00424 **
## ms_dx_condensedMS, Subtype Not Specified 0.54156
## race_ethnicity_cleanOther/Unknown/Declined 0.31123
## clean_sexMale 0.08351 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2441 on 23 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.7127, Adjusted R-squared: 0.6253
## F-statistic: 8.151 on 7 and 23 DF, p-value: 5.259e-05
metric_regression(home_df, t25fw_log, foot1_tot_double_support_time_mean_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 32 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.48968 -0.23291 -0.01979 0.20780 0.57054
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.2472 0.0880 14.173 1.44e-14
## foot1_tot_double_support_time_mean_pose_hv 0.8724 0.1495 5.837 2.50e-06
##
## (Intercept) ***
## foot1_tot_double_support_time_mean_pose_hv ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.275 on 29 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.5402, Adjusted R-squared: 0.5243
## F-statistic: 34.07 on 1 and 29 DF, p-value: 2.496e-06
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.34331 -0.14480 -0.00802 0.12458 0.49292
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.182544 0.264666 4.468
## foot1_tot_double_support_time_mean_pose_hv 0.491971 0.223831 2.198
## demoEHR_Age 0.003794 0.005056 0.750
## demoEHR_DiseaseDuration 0.005721 0.008742 0.654
## ms_dx_condensedProgressive MS 0.362727 0.217019 1.671
## ms_dx_condensedMS, Subtype Not Specified -0.110090 0.167710 -0.656
## race_ethnicity_cleanOther/Unknown/Declined -0.175103 0.168754 -1.038
## clean_sexMale -0.227590 0.134525 -1.692
## Pr(>|t|)
## (Intercept) 0.000175 ***
## foot1_tot_double_support_time_mean_pose_hv 0.038284 *
## demoEHR_Age 0.460577
## demoEHR_DiseaseDuration 0.519332
## ms_dx_condensedProgressive MS 0.108194
## ms_dx_condensedMS, Subtype Not Specified 0.518063
## race_ethnicity_cleanOther/Unknown/Declined 0.310230
## clean_sexMale 0.104189
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2358 on 23 degrees of freedom
## (32 observations deleted due to missingness)
## Multiple R-squared: 0.7318, Adjusted R-squared: 0.6502
## F-statistic: 8.966 on 7 and 23 DF, p-value: 2.527e-05
# home left
metric_regression(home_l_df, t25fw_log, foot1_tot_double_support_time_mean_pose_hv)## [1] "Data Frame: home_l_df"## Warning: Removed 18 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51134 -0.24628 0.06383 0.15994 0.57887
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.2089 0.1582 7.643 1.01e-05
## foot1_tot_double_support_time_mean_pose_hv 0.9723 0.3024 3.216 0.00822
##
## (Intercept) ***
## foot1_tot_double_support_time_mean_pose_hv **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2985 on 11 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.4846, Adjusted R-squared: 0.4377
## F-statistic: 10.34 on 1 and 11 DF, p-value: 0.00822
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## 1 9 13 15 17 22 28
## -1.271e-01 -1.132e-01 1.753e-03 1.310e-01 1.271e-01 -1.735e-17 -6.903e-02
## 38 42 52 56 60 62
## 1.132e-01 -1.671e-01 -1.102e-01 4.133e-01 -1.996e-01 -3.469e-18
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 1.398802 0.381377 3.668
## foot1_tot_double_support_time_mean_pose_hv 0.784056 0.422640 1.855
## demoEHR_Age 0.004114 0.008431 0.488
## demoEHR_DiseaseDuration -0.021115 0.021555 -0.980
## ms_dx_condensedProgressive MS 0.068055 0.345660 0.197
## ms_dx_condensedMS, Subtype Not Specified -0.384447 0.318054 -1.209
## race_ethnicity_cleanOther/Unknown/Declined -0.291739 0.292978 -0.996
## clean_sexMale -0.611205 0.258827 -2.361
## Pr(>|t|)
## (Intercept) 0.0145 *
## foot1_tot_double_support_time_mean_pose_hv 0.1227
## demoEHR_Age 0.6462
## demoEHR_DiseaseDuration 0.3723
## ms_dx_condensedProgressive MS 0.8517
## ms_dx_condensedMS, Subtype Not Specified 0.2808
## race_ethnicity_cleanOther/Unknown/Declined 0.3651
## clean_sexMale 0.0646 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2571 on 5 degrees of freedom
## (18 observations deleted due to missingness)
## Multiple R-squared: 0.8261, Adjusted R-squared: 0.5828
## F-statistic: 3.394 on 7 and 5 DF, p-value: 0.09862
# home right
metric_regression(home_r_df, t25fw_log, foot1_tot_double_support_time_mean_pose_hv)## [1] "Data Frame: home_r_df"## Warning: Removed 14 rows containing missing values (`geom_point()`).
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.45459 -0.20532 -0.01809 0.20460 0.46910
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.2627 0.1123 11.242 5.26e-09
## foot1_tot_double_support_time_mean_pose_hv 0.8337 0.1777 4.691 0.000245
##
## (Intercept) ***
## foot1_tot_double_support_time_mean_pose_hv ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2737 on 16 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.579, Adjusted R-squared: 0.5527
## F-statistic: 22 on 1 and 16 DF, p-value: 0.0002454
## [1] "t25fw_log ~ foot1_tot_double_support_time_mean_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.37020 -0.14211 0.01509 0.13875 0.42195
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.7286486 0.5620906 1.296
## foot1_tot_double_support_time_mean_pose_hv 0.6897245 0.4137712 1.667
## demoEHR_Age 0.0119207 0.0103643 1.150
## demoEHR_DiseaseDuration -0.0002769 0.0139011 -0.020
## ms_dx_condensedProgressive MS 0.1703523 0.4274553 0.399
## ms_dx_condensedMS, Subtype Not Specified -0.1901960 0.2704161 -0.703
## race_ethnicity_cleanOther/Unknown/Declined 0.0022200 0.2613434 0.008
## clean_sexMale -0.0154889 0.2438597 -0.064
## Pr(>|t|)
## (Intercept) 0.224
## foot1_tot_double_support_time_mean_pose_hv 0.126
## demoEHR_Age 0.277
## demoEHR_DiseaseDuration 0.984
## ms_dx_condensedProgressive MS 0.699
## ms_dx_condensedMS, Subtype Not Specified 0.498
## race_ethnicity_cleanOther/Unknown/Declined 0.993
## clean_sexMale 0.951
##
## Residual standard error: 0.2572 on 10 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7677, Adjusted R-squared: 0.6051
## F-statistic: 4.722 on 7 and 10 DF, p-value: 0.01394
Multivariate - home, Gait metrics only –> T25fw
Metrics only - not including double support/stance measures, too many missing. May include after improving code
# Home
# confounding +
home_t25fw_multivar_model <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv,
data = home_df)
summary(home_t25fw_multivar_model)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.38964 -0.17589 0.06283 0.18239 0.46609
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.473285 0.461582 1.025 0.3108
## log_delta_pix_h_rel_median_pose_hv -0.290997 0.118489 -2.456 0.0181 *
## stride_time_median_sec_pose_hv 0.467391 0.224108 2.086 0.0429 *
## mean_cadence_step_per_min_pose_hv -0.003097 0.002533 -1.223 0.2280
## stride_width_median_cm_pose_hv 0.045898 0.015859 2.894 0.0059 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2345 on 44 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.6534, Adjusted R-squared: 0.6219
## F-statistic: 20.74 on 4 and 44 DF, p-value: 1.154e-09hist(resid(home_t25fw_multivar_model))
# interaction *
home_t25fw_multivar_model_2 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv *
stride_time_median_sec_pose_hv *
mean_cadence_step_per_min_pose_hv *
stride_width_median_cm_pose_hv,
data = home_df)
summary(home_t25fw_multivar_model_2)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv *
## stride_time_median_sec_pose_hv * mean_cadence_step_per_min_pose_hv *
## stride_width_median_cm_pose_hv, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.34514 -0.14413 -0.00139 0.15571 0.31573
##
## Coefficients:
## Estimate
## (Intercept) 27.22718
## log_delta_pix_h_rel_median_pose_hv 29.31956
## stride_time_median_sec_pose_hv -20.32178
## mean_cadence_step_per_min_pose_hv -0.32198
## stride_width_median_cm_pose_hv -2.33774
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv -22.57757
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv -0.33906
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.25505
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv -2.32492
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.94475
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.03053
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.26528
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.81019
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.02815
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.02552
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.02253
## Std. Error
## (Intercept) 33.35425
## log_delta_pix_h_rel_median_pose_hv 28.84964
## stride_time_median_sec_pose_hv 29.78914
## mean_cadence_step_per_min_pose_hv 0.39228
## stride_width_median_cm_pose_hv 2.16025
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv 23.85350
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv 0.34139
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.36598
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv 1.85602
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.92554
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.02593
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.29876
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.49564
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.02261
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.02437
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.01963
## t value
## (Intercept) 0.816
## log_delta_pix_h_rel_median_pose_hv 1.016
## stride_time_median_sec_pose_hv -0.682
## mean_cadence_step_per_min_pose_hv -0.821
## stride_width_median_cm_pose_hv -1.082
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv -0.947
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv -0.993
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.697
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv -1.253
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.010
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 1.178
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.888
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 1.210
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 1.245
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -1.047
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -1.148
## Pr(>|t|)
## (Intercept) 0.420
## log_delta_pix_h_rel_median_pose_hv 0.317
## stride_time_median_sec_pose_hv 0.500
## mean_cadence_step_per_min_pose_hv 0.418
## stride_width_median_cm_pose_hv 0.287
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv 0.351
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv 0.328
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.491
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv 0.219
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 0.320
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.247
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.381
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 0.235
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.222
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.303
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.259
##
## Residual standard error: 0.2068 on 33 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7979, Adjusted R-squared: 0.706
## F-statistic: 8.685 on 15 and 33 DF, p-value: 1.427e-07hist(resid(home_t25fw_multivar_model_2))
Multivariate - home, Gait metrics + disease + demographics –> T25fw
# Home
# Metrics + disease and demographic info
# add MS subtype
home_t25fw_multivar_model_3 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed,
data = home_df)
summary(home_t25fw_multivar_model_3)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.39678 -0.20485 0.02895 0.16556 0.33456
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.949884 0.478360 1.986 0.05362
## log_delta_pix_h_rel_median_pose_hv -0.173257 0.120180 -1.442 0.15682
## stride_time_median_sec_pose_hv -0.019184 0.284086 -0.068 0.94648
## mean_cadence_step_per_min_pose_hv -0.002312 0.002393 -0.966 0.33947
## stride_width_median_cm_pose_hv 0.053633 0.015880 3.377 0.00159
## ms_dx_condensedProgressive MS 0.529568 0.203319 2.605 0.01266
## ms_dx_condensedMS, Subtype Not Specified -0.098729 0.126304 -0.782 0.43879
##
## (Intercept) .
## log_delta_pix_h_rel_median_pose_hv
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv **
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2199 on 42 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7089, Adjusted R-squared: 0.6673
## F-statistic: 17.05 on 6 and 42 DF, p-value: 7.328e-10hist(resid(home_t25fw_multivar_model_3))
# add age
home_t25fw_multivar_model_4 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age,
data = home_df)
summary(home_t25fw_multivar_model_4)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed + demoEHR_Age,
## data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.40818 -0.14961 0.04824 0.12476 0.39527
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.484495 0.453209 1.069 0.29131
## log_delta_pix_h_rel_median_pose_hv -0.107680 0.109998 -0.979 0.33336
## stride_time_median_sec_pose_hv 0.234346 0.267065 0.877 0.38533
## mean_cadence_step_per_min_pose_hv -0.002027 0.002155 -0.940 0.35260
## stride_width_median_cm_pose_hv 0.040708 0.014824 2.746 0.00892
## ms_dx_condensedProgressive MS 0.388637 0.187957 2.068 0.04502
## ms_dx_condensedMS, Subtype Not Specified -0.242951 0.121841 -1.994 0.05283
## demoEHR_Age 0.007966 0.002420 3.292 0.00205
##
## (Intercept)
## log_delta_pix_h_rel_median_pose_hv
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv **
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified .
## demoEHR_Age **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.198 on 41 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7698, Adjusted R-squared: 0.7305
## F-statistic: 19.58 on 7 and 41 DF, p-value: 3.176e-11hist(resid(home_t25fw_multivar_model_4))
# add disease duration
home_t25fw_multivar_model_5 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration,
data = home_df)
summary(home_t25fw_multivar_model_5)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.39030 -0.15718 0.02653 0.12222 0.38935
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.445364 0.457749 0.973 0.33643
## log_delta_pix_h_rel_median_pose_hv -0.112329 0.110623 -1.015 0.31601
## stride_time_median_sec_pose_hv 0.255871 0.269545 0.949 0.34818
## mean_cadence_step_per_min_pose_hv -0.001880 0.002172 -0.865 0.39203
## stride_width_median_cm_pose_hv 0.040532 0.014890 2.722 0.00956
## ms_dx_condensedProgressive MS 0.394215 0.188894 2.087 0.04331
## ms_dx_condensedMS, Subtype Not Specified -0.249582 0.122643 -2.035 0.04851
## demoEHR_Age 0.007115 0.002650 2.685 0.01051
## demoEHR_DiseaseDuration 0.003405 0.004228 0.805 0.42538
##
## (Intercept)
## log_delta_pix_h_rel_median_pose_hv
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv **
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified *
## demoEHR_Age *
## demoEHR_DiseaseDuration
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1988 on 40 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7734, Adjusted R-squared: 0.7281
## F-statistic: 17.07 on 8 and 40 DF, p-value: 1.085e-10hist(resid(home_t25fw_multivar_model_5))
# add race and ethnicity
home_t25fw_multivar_model_6 <- lm(t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = home_df)
summary(home_t25fw_multivar_model_6)##
## Call:
## lm(formula = t25fw_log ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + stride_width_median_cm_pose_hv +
## ms_dx_condensed + demoEHR_Age + demoEHR_DiseaseDuration +
## race_ethnicity_clean, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.36318 -0.11606 0.01794 0.10324 0.37890
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 0.065546 0.326814 0.201
## log_delta_pix_h_rel_median_pose_hv -0.122401 0.108846 -1.125
## stride_time_median_sec_pose_hv 0.342211 0.246698 1.387
## stride_width_median_cm_pose_hv 0.038408 0.014878 2.582
## ms_dx_condensedProgressive MS 0.383923 0.185693 2.068
## ms_dx_condensedMS, Subtype Not Specified -0.252643 0.126893 -1.991
## demoEHR_Age 0.004604 0.002898 1.589
## demoEHR_DiseaseDuration 0.005998 0.004857 1.235
## race_ethnicity_cleanHispanic or Latino 0.236485 0.228765 1.034
## race_ethnicity_cleanWhite Not Hispanic 0.235283 0.150759 1.561
## race_ethnicity_cleanOther/Unknown/Declined 0.096786 0.161344 0.600
## Pr(>|t|)
## (Intercept) 0.8421
## log_delta_pix_h_rel_median_pose_hv 0.2678
## stride_time_median_sec_pose_hv 0.1735
## stride_width_median_cm_pose_hv 0.0138 *
## ms_dx_condensedProgressive MS 0.0455 *
## ms_dx_condensedMS, Subtype Not Specified 0.0537 .
## demoEHR_Age 0.1204
## demoEHR_DiseaseDuration 0.2245
## race_ethnicity_cleanHispanic or Latino 0.3078
## race_ethnicity_cleanWhite Not Hispanic 0.1269
## race_ethnicity_cleanOther/Unknown/Declined 0.5522
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1952 on 38 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7926, Adjusted R-squared: 0.7381
## F-statistic: 14.53 on 10 and 38 DF, p-value: 3.836e-10hist(resid(home_t25fw_multivar_model_6))
Linear Regression –> PWS Velocity Zeno Velocity
Preferred walking speed
Demographics Only
pws_matvel_dem_model <- lm(PWS_velocitycmsecmean ~
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = zeno_pws_df)
summary(pws_matvel_dem_model)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -64.277 -16.815 2.337 18.755 69.779
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 120.29915 9.50111 12.662
## ms_dx_condensedProgressive MS -33.69399 5.36047 -6.286
## ms_dx_condensedMS, Subtype Not Specified 13.96842 19.26563 0.725
## demoEHR_Age -0.10435 0.19632 -0.532
## demoEHR_DiseaseDuration 0.01945 0.25790 0.075
## race_ethnicity_cleanBlack Or African American -26.52248 9.83961 -2.695
## race_ethnicity_cleanHispanic or Latino -6.14212 8.82214 -0.696
## race_ethnicity_cleanWhite Not Hispanic -5.00506 7.21961 -0.693
## race_ethnicity_cleanOther/Unknown/Declined -6.10194 9.05229 -0.674
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## ms_dx_condensedProgressive MS 1.89e-09 ***
## ms_dx_condensedMS, Subtype Not Specified 0.4692
## demoEHR_Age 0.5956
## demoEHR_DiseaseDuration 0.9400
## race_ethnicity_cleanBlack Or African American 0.0076 **
## race_ethnicity_cleanHispanic or Latino 0.4871
## race_ethnicity_cleanWhite Not Hispanic 0.4889
## race_ethnicity_cleanOther/Unknown/Declined 0.5010
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.65 on 208 degrees of freedom
## Multiple R-squared: 0.239, Adjusted R-squared: 0.2097
## F-statistic: 8.166 on 8 and 208 DF, p-value: 1.369e-09hist(resid(pws_matvel_dem_model))
pws_matvel_dem_model_2 <- lm(PWS_velocitycmsecmean ~
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = zeno_pws_uniqueid_df)
summary(pws_matvel_dem_model_2)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = zeno_pws_uniqueid_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.104 -17.662 -0.462 18.947 62.095
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 118.66079 11.33605 10.468
## ms_dx_condensedProgressive MS -35.37281 6.14574 -5.756
## ms_dx_condensedMS, Subtype Not Specified 18.00988 19.36443 0.930
## demoEHR_Age -0.03415 0.21723 -0.157
## demoEHR_DiseaseDuration -0.23638 0.32038 -0.738
## race_ethnicity_cleanBlack Or African American -20.94183 11.49151 -1.822
## race_ethnicity_cleanHispanic or Latino -7.56336 10.58983 -0.714
## race_ethnicity_cleanWhite Not Hispanic -7.54175 8.75675 -0.861
## race_ethnicity_cleanOther/Unknown/Declined -7.56229 11.16546 -0.677
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## ms_dx_condensedProgressive MS 5.06e-08 ***
## ms_dx_condensedMS, Subtype Not Specified 0.3539
## demoEHR_Age 0.8753
## demoEHR_DiseaseDuration 0.4618
## race_ethnicity_cleanBlack Or African American 0.0705 .
## race_ethnicity_cleanHispanic or Latino 0.4763
## race_ethnicity_cleanWhite Not Hispanic 0.3905
## race_ethnicity_cleanOther/Unknown/Declined 0.4993
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.61 on 143 degrees of freedom
## Multiple R-squared: 0.2573, Adjusted R-squared: 0.2158
## F-statistic: 6.194 on 8 and 143 DF, p-value: 7.358e-07hist(resid(pws_matvel_dem_model_2))
# Preferred walking speed Zeno
hist(zeno_pws_df$PWS_velocitycmsecmean)
ggplot(data = zeno_pws_df, aes(x = log_delta_pix_h_rel_median_pose_zv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 14 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_time_median_sec_pose_zv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 48 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = mean_cadence_step_per_min_pose_zv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
ggplot(data = zeno_pws_df, aes(x = stride_width_median_cm_pose_zv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 40 rows containing missing values (`geom_point()`).
Univariate regressions - each video metric
metric_regression(zeno_pws_df, PWS_velocitycmsecmean, log_delta_pix_h_rel_median_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 14 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -60.968 -20.258 1.462 14.689 85.410
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 146.342 5.449 26.856 < 2e-16 ***
## log_delta_pix_h_rel_median_pose_zv 30.650 3.631 8.441 6.21e-15 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.44 on 201 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.2617, Adjusted R-squared: 0.258
## F-statistic: 71.24 on 1 and 201 DF, p-value: 6.212e-15
## [1] "PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -57.20 -15.56 0.47 17.50 68.71
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 160.966620 11.023121 14.603
## log_delta_pix_h_rel_median_pose_zv 23.490487 3.785996 6.205
## demoEHR_Age -0.200511 0.184141 -1.089
## demoEHR_DiseaseDuration 0.002313 0.243415 0.010
## ms_dx_condensedProgressive MS -22.137904 5.651293 -3.917
## ms_dx_condensedMS, Subtype Not Specified 11.658966 17.855576 0.653
## race_ethnicity_cleanBlack Or African American -31.210375 9.360920 -3.334
## race_ethnicity_cleanHispanic or Latino -12.462700 8.691511 -1.434
## race_ethnicity_cleanWhite Not Hispanic -9.399209 7.050092 -1.333
## race_ethnicity_cleanOther/Unknown/Declined -11.009483 8.830550 -1.247
## clean_sexMale -2.243003 4.236191 -0.529
## clean_sexNon-Binary -2.580832 24.874815 -0.104
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## log_delta_pix_h_rel_median_pose_zv 3.33e-09 ***
## demoEHR_Age 0.277571
## demoEHR_DiseaseDuration 0.992430
## ms_dx_condensedProgressive MS 0.000125 ***
## ms_dx_condensedMS, Subtype Not Specified 0.514568
## race_ethnicity_cleanBlack Or African American 0.001028 **
## race_ethnicity_cleanHispanic or Latino 0.153238
## race_ethnicity_cleanWhite Not Hispanic 0.184053
## race_ethnicity_cleanOther/Unknown/Declined 0.214016
## clean_sexMale 0.597083
## clean_sexNon-Binary 0.917474
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 24.61 on 191 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.3921, Adjusted R-squared: 0.3571
## F-statistic: 11.2 on 11 and 191 DF, p-value: 6.393e-16
metric_regression(zeno_pws_df, PWS_velocitycmsecmean, stride_time_median_sec_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 48 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ stride_time_median_sec_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.262 -14.416 1.544 16.620 49.338
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 210.13 13.45 15.620 < 2e-16 ***
## stride_time_median_sec_pose_zv -94.27 11.87 -7.939 2.84e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 22.78 on 167 degrees of freedom
## (48 observations deleted due to missingness)
## Multiple R-squared: 0.274, Adjusted R-squared: 0.2696
## F-statistic: 63.02 on 1 and 167 DF, p-value: 2.84e-13
## [1] "PWS_velocitycmsecmean ~ stride_time_median_sec_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -94.695 -13.909 2.375 14.538 44.584
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 213.5351 16.6498 12.825
## stride_time_median_sec_pose_zv -85.9863 11.9821 -7.176
## demoEHR_Age -0.1684 0.1851 -0.910
## demoEHR_DiseaseDuration -0.3027 0.2486 -1.218
## ms_dx_condensedProgressive MS -12.1565 5.4398 -2.235
## ms_dx_condensedMS, Subtype Not Specified 15.8514 15.7677 1.005
## race_ethnicity_cleanBlack Or African American -20.1646 9.0572 -2.226
## race_ethnicity_cleanHispanic or Latino 6.4141 8.2043 0.782
## race_ethnicity_cleanWhite Not Hispanic 1.1599 6.8861 0.168
## race_ethnicity_cleanOther/Unknown/Declined -3.6449 9.6685 -0.377
## clean_sexMale 1.1758 3.9167 0.300
## clean_sexNon-Binary 1.9397 21.9132 0.089
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## stride_time_median_sec_pose_zv 2.68e-11 ***
## demoEHR_Age 0.3644
## demoEHR_DiseaseDuration 0.2251
## ms_dx_condensedProgressive MS 0.0268 *
## ms_dx_condensedMS, Subtype Not Specified 0.3163
## race_ethnicity_cleanBlack Or African American 0.0274 *
## race_ethnicity_cleanHispanic or Latino 0.4355
## race_ethnicity_cleanWhite Not Hispanic 0.8665
## race_ethnicity_cleanOther/Unknown/Declined 0.7067
## clean_sexMale 0.7644
## clean_sexNon-Binary 0.9296
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.67 on 157 degrees of freedom
## (48 observations deleted due to missingness)
## Multiple R-squared: 0.3825, Adjusted R-squared: 0.3392
## F-statistic: 8.84 on 11 and 157 DF, p-value: 4.017e-12
metric_regression(zeno_pws_df, PWS_velocitycmsecmean, mean_cadence_step_per_min_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 40 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ mean_cadence_step_per_min_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -72.074 -14.455 2.047 15.820 49.555
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.9456 12.2997 -0.321 0.749
## mean_cadence_step_per_min_pose_zv 1.0332 0.1176 8.785 1.41e-15 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 23.27 on 175 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.306, Adjusted R-squared: 0.3021
## F-statistic: 77.17 on 1 and 175 DF, p-value: 1.411e-15
## [1] "PWS_velocitycmsecmean ~ mean_cadence_step_per_min_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -73.636 -15.400 2.225 13.757 45.370
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 13.6417 15.3470 0.889
## mean_cadence_step_per_min_pose_zv 0.9643 0.1168 8.257
## demoEHR_Age -0.1830 0.1784 -1.026
## demoEHR_DiseaseDuration -0.2379 0.2376 -1.001
## ms_dx_condensedProgressive MS -16.9654 5.0872 -3.335
## ms_dx_condensedMS, Subtype Not Specified 10.6654 15.6198 0.683
## race_ethnicity_cleanBlack Or African American -20.8631 8.9404 -2.334
## race_ethnicity_cleanHispanic or Latino 3.5656 8.0369 0.444
## race_ethnicity_cleanWhite Not Hispanic 4.2542 6.7972 0.626
## race_ethnicity_cleanOther/Unknown/Declined 6.5356 9.4753 0.690
## clean_sexMale 4.8437 3.8322 1.264
## clean_sexNon-Binary -2.0406 21.7045 -0.094
## Pr(>|t|)
## (Intercept) 0.37536
## mean_cadence_step_per_min_pose_zv 4.6e-14 ***
## demoEHR_Age 0.30645
## demoEHR_DiseaseDuration 0.31820
## ms_dx_condensedProgressive MS 0.00105 **
## ms_dx_condensedMS, Subtype Not Specified 0.49568
## race_ethnicity_cleanBlack Or African American 0.02082 *
## race_ethnicity_cleanHispanic or Latino 0.65788
## race_ethnicity_cleanWhite Not Hispanic 0.53226
## race_ethnicity_cleanOther/Unknown/Declined 0.49132
## clean_sexMale 0.20803
## clean_sexNon-Binary 0.92521
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.45 on 165 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.4437, Adjusted R-squared: 0.4066
## F-statistic: 11.96 on 11 and 165 DF, p-value: 2.508e-16
metric_regression(zeno_pws_df, PWS_velocitycmsecmean, stride_width_median_cm_pose_zv)## [1] "Data Frame: zeno_pws_df"## Warning: Removed 40 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ stride_width_median_cm_pose_zv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -101.589 -15.870 5.143 19.832 71.560
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 129.1508 7.7260 16.716 < 2e-16 ***
## stride_width_median_cm_pose_zv -2.0966 0.5978 -3.507 0.000575 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 27 on 175 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.06568, Adjusted R-squared: 0.06034
## F-statistic: 12.3 on 1 and 175 DF, p-value: 0.0005749
## [1] "PWS_velocitycmsecmean ~ stride_width_median_cm_pose_zv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -73.605 -17.520 4.266 19.684 63.927
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 131.13849 13.18956 9.943
## stride_width_median_cm_pose_zv -1.10097 0.60153 -1.830
## demoEHR_Age -0.18336 0.20999 -0.873
## demoEHR_DiseaseDuration 0.03058 0.27700 0.110
## ms_dx_condensedProgressive MS -25.33483 6.00672 -4.218
## ms_dx_condensedMS, Subtype Not Specified 17.85317 18.40265 0.970
## race_ethnicity_cleanBlack Or African American -24.71917 10.51202 -2.352
## race_ethnicity_cleanHispanic or Latino 0.40312 9.54618 0.042
## race_ethnicity_cleanWhite Not Hispanic -0.36837 8.03850 -0.046
## race_ethnicity_cleanOther/Unknown/Declined -10.93036 10.96116 -0.997
## clean_sexMale 2.03036 4.50022 0.451
## clean_sexNon-Binary 7.69159 25.52853 0.301
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## stride_width_median_cm_pose_zv 0.0690 .
## demoEHR_Age 0.3838
## demoEHR_DiseaseDuration 0.9122
## ms_dx_condensedProgressive MS 4.06e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.3334
## race_ethnicity_cleanBlack Or African American 0.0199 *
## race_ethnicity_cleanHispanic or Latino 0.9664
## race_ethnicity_cleanWhite Not Hispanic 0.9635
## race_ethnicity_cleanOther/Unknown/Declined 0.3201
## clean_sexMale 0.6525
## clean_sexNon-Binary 0.7636
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 25.25 on 165 degrees of freedom
## (40 observations deleted due to missingness)
## Multiple R-squared: 0.2295, Adjusted R-squared: 0.1781
## F-statistic: 4.468 on 11 and 165 DF, p-value: 6.822e-06
Multivariate regression
# Metrics only - not including double support/stance measures, too many missing
# confounding +
pws_matvel_multivar_model <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv,
data = zeno_pws_df)
summary(pws_matvel_multivar_model)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.54 -13.18 0.82 16.24 40.18
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 101.9025 36.6945 2.777 0.006176 **
## log_delta_pix_h_rel_median_pose_zv 4.0642 4.1088 0.989 0.324163
## stride_time_median_sec_pose_zv -42.1065 16.4597 -2.558 0.011501 *
## mean_cadence_step_per_min_pose_zv 0.6940 0.1945 3.568 0.000481 ***
## stride_width_median_cm_pose_zv -1.4094 0.5540 -2.544 0.011948 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.81 on 152 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.3782, Adjusted R-squared: 0.3619
## F-statistic: 23.11 on 4 and 152 DF, p-value: 6.166e-15hist(resid(pws_matvel_multivar_model))
# PWS interaction *
pws_matvel_multivar_model_2 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv *
stride_time_median_sec_pose_zv *
mean_cadence_step_per_min_pose_zv *
stride_width_median_cm_pose_zv,
data = zeno_pws_df)
summary(pws_matvel_multivar_model_2)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv *
## stride_time_median_sec_pose_zv * mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -73.827 -11.254 0.822 14.280 36.320
##
## Coefficients:
## Estimate
## (Intercept) 2318.8462
## log_delta_pix_h_rel_median_pose_zv 883.5118
## stride_time_median_sec_pose_zv -1747.2510
## mean_cadence_step_per_min_pose_zv -18.9828
## stride_width_median_cm_pose_zv -177.1575
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv -570.6399
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv -7.2171
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 14.8183
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv -72.9952
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 141.7145
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 1.6658
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 4.2169
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 52.6956
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.6679
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -1.3345
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -0.4658
## Std. Error
## (Intercept) 1115.9297
## log_delta_pix_h_rel_median_pose_zv 784.9593
## stride_time_median_sec_pose_zv 858.4759
## mean_cadence_step_per_min_pose_zv 9.4037
## stride_width_median_cm_pose_zv 89.1389
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 529.7408
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 7.2291
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 7.3381
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 58.4011
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 72.6931
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.7697
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 5.1632
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 39.8877
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.5388
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.6395
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.3814
## t value
## (Intercept) 2.078
## log_delta_pix_h_rel_median_pose_zv 1.126
## stride_time_median_sec_pose_zv -2.035
## mean_cadence_step_per_min_pose_zv -2.019
## stride_width_median_cm_pose_zv -1.987
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv -1.077
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv -0.998
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 2.019
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv -1.250
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 1.949
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 2.164
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.817
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 1.321
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 1.240
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -2.087
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv -1.221
## Pr(>|t|)
## (Intercept) 0.0395
## log_delta_pix_h_rel_median_pose_zv 0.2623
## stride_time_median_sec_pose_zv 0.0437
## mean_cadence_step_per_min_pose_zv 0.0454
## stride_width_median_cm_pose_zv 0.0488
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv 0.2832
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv 0.3198
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.0453
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv 0.2134
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.0532
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.0321
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv 0.4155
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv 0.1886
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.2172
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.0387
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv 0.2240
##
## (Intercept) *
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv *
## stride_width_median_cm_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv .
## mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:stride_width_median_cm_pose_zv
## log_delta_pix_h_rel_median_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv *
## log_delta_pix_h_rel_median_pose_zv:stride_time_median_sec_pose_zv:mean_cadence_step_per_min_pose_zv:stride_width_median_cm_pose_zv
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.04 on 141 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.5129, Adjusted R-squared: 0.4611
## F-statistic: 9.898 on 15 and 141 DF, p-value: 1.041e-15hist(resid(pws_matvel_multivar_model_2))
# PWS
# Metrics + disease and demographic info
# add MS subtype
pws_matvel_multivar_model_3 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed,
data = zeno_pws_df)
summary(pws_matvel_multivar_model_3)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -92.261 -13.998 1.211 15.668 41.027
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 87.4055 36.8374 2.373 0.018925
## log_delta_pix_h_rel_median_pose_zv 1.9669 4.1449 0.475 0.635805
## stride_time_median_sec_pose_zv -36.4959 16.4354 -2.221 0.027879
## mean_cadence_step_per_min_pose_zv 0.7136 0.1925 3.706 0.000295
## stride_width_median_cm_pose_zv -1.0277 0.5762 -1.783 0.076529
## ms_dx_condensedProgressive MS -13.4480 5.7845 -2.325 0.021422
## ms_dx_condensedMS, Subtype Not Specified 10.4326 15.4306 0.676 0.500021
##
## (Intercept) *
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv ***
## stride_width_median_cm_pose_zv .
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.53 on 150 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4022, Adjusted R-squared: 0.3783
## F-statistic: 16.82 on 6 and 150 DF, p-value: 8.603e-15hist(resid(pws_matvel_multivar_model_3))
# add age
pws_matvel_multivar_model_4 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age,
data = zeno_pws_df)
summary(pws_matvel_multivar_model_4)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age,
## data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -92.606 -12.226 2.035 14.588 41.748
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 107.5345 37.4285 2.873 0.004659
## log_delta_pix_h_rel_median_pose_zv 2.8095 4.1067 0.684 0.494954
## stride_time_median_sec_pose_zv -39.2791 16.2634 -2.415 0.016939
## mean_cadence_step_per_min_pose_zv 0.7174 0.1900 3.776 0.000229
## stride_width_median_cm_pose_zv -1.0270 0.5686 -1.806 0.072899
## ms_dx_condensedProgressive MS -8.9931 6.0402 -1.489 0.138634
## ms_dx_condensedMS, Subtype Not Specified 16.7838 15.4838 1.084 0.280135
## demoEHR_Age -0.3410 0.1513 -2.253 0.025705
##
## (Intercept) **
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv ***
## stride_width_median_cm_pose_zv .
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.24 on 149 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4219, Adjusted R-squared: 0.3947
## F-statistic: 15.53 on 7 and 149 DF, p-value: 3.443e-15hist(resid(pws_matvel_multivar_model_4))
# add disease duration
pws_matvel_multivar_model_5 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration,
data = zeno_pws_df)
summary(pws_matvel_multivar_model_5)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -95.797 -12.299 1.693 14.640 41.342
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 105.2216 37.4396 2.810 0.005617
## log_delta_pix_h_rel_median_pose_zv 2.4070 4.1168 0.585 0.559652
## stride_time_median_sec_pose_zv -40.9970 16.3128 -2.513 0.013036
## mean_cadence_step_per_min_pose_zv 0.7289 0.1900 3.836 0.000185
## stride_width_median_cm_pose_zv -0.9820 0.5693 -1.725 0.086619
## ms_dx_condensedProgressive MS -8.4398 6.0523 -1.394 0.165263
## ms_dx_condensedMS, Subtype Not Specified 17.1799 15.4701 1.111 0.268574
## demoEHR_Age -0.2439 0.1729 -1.410 0.160633
## demoEHR_DiseaseDuration -0.2870 0.2481 -1.157 0.249262
##
## (Intercept) **
## log_delta_pix_h_rel_median_pose_zv
## stride_time_median_sec_pose_zv *
## mean_cadence_step_per_min_pose_zv ***
## stride_width_median_cm_pose_zv .
## ms_dx_condensedProgressive MS
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age
## demoEHR_DiseaseDuration
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.22 on 148 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4271, Adjusted R-squared: 0.3961
## F-statistic: 13.79 on 8 and 148 DF, p-value: 7.57e-15hist(resid(pws_matvel_multivar_model_5))
# add race and ethnicity
pws_matvel_multivar_model_6 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
stride_time_median_sec_pose_zv +
mean_cadence_step_per_min_pose_zv +
stride_width_median_cm_pose_zv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = zeno_pws_df)
summary(pws_matvel_multivar_model_6)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_zv +
## stride_time_median_sec_pose_zv + mean_cadence_step_per_min_pose_zv +
## stride_width_median_cm_pose_zv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = zeno_pws_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -92.842 -11.784 2.359 12.628 40.435
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 91.9506 38.3256 2.399
## log_delta_pix_h_rel_median_pose_zv 2.4915 4.1081 0.606
## stride_time_median_sec_pose_zv -35.5615 16.0590 -2.214
## mean_cadence_step_per_min_pose_zv 0.7586 0.1892 4.009
## stride_width_median_cm_pose_zv -0.6119 0.5685 -1.076
## ms_dx_condensedProgressive MS -11.0238 5.9488 -1.853
## ms_dx_condensedMS, Subtype Not Specified 14.5049 15.0923 0.961
## demoEHR_Age -0.2160 0.1799 -1.201
## demoEHR_DiseaseDuration -0.2742 0.2420 -1.133
## race_ethnicity_cleanBlack Or African American -21.6155 9.1324 -2.367
## race_ethnicity_cleanHispanic or Latino 3.5964 8.6421 0.416
## race_ethnicity_cleanWhite Not Hispanic -0.3255 7.2194 -0.045
## race_ethnicity_cleanOther/Unknown/Declined -1.1066 9.7475 -0.114
## Pr(>|t|)
## (Intercept) 0.0177 *
## log_delta_pix_h_rel_median_pose_zv 0.5452
## stride_time_median_sec_pose_zv 0.0284 *
## mean_cadence_step_per_min_pose_zv 9.75e-05 ***
## stride_width_median_cm_pose_zv 0.2836
## ms_dx_condensedProgressive MS 0.0659 .
## ms_dx_condensedMS, Subtype Not Specified 0.3381
## demoEHR_Age 0.2319
## demoEHR_DiseaseDuration 0.2591
## race_ethnicity_cleanBlack Or African American 0.0193 *
## race_ethnicity_cleanHispanic or Latino 0.6779
## race_ethnicity_cleanWhite Not Hispanic 0.9641
## race_ethnicity_cleanOther/Unknown/Declined 0.9098
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.66 on 144 degrees of freedom
## (60 observations deleted due to missingness)
## Multiple R-squared: 0.4713, Adjusted R-squared: 0.4272
## F-statistic: 10.7 on 12 and 144 DF, p-value: 5.858e-15hist(resid(pws_matvel_multivar_model_6))
Linear Regression –> PWS Velocity Home Video
Demographics Only
home_matvel_dem_model <- lm(PWS_velocitycmsecmean ~
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = home_df)
summary(home_matvel_dem_model)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -37.53 -13.44 0.00 16.36 63.21
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 132.2485 15.7702 8.386
## ms_dx_condensedProgressive MS -43.5841 9.3090 -4.682
## ms_dx_condensedMS, Subtype Not Specified 17.3052 12.5899 1.375
## demoEHR_Age -0.1137 0.2730 -0.417
## demoEHR_DiseaseDuration -0.3132 0.4720 -0.664
## race_ethnicity_cleanBlack Or African American -72.3140 20.8488 -3.469
## race_ethnicity_cleanHispanic or Latino -31.1139 21.7722 -1.429
## race_ethnicity_cleanWhite Not Hispanic -13.6725 11.9575 -1.143
## race_ethnicity_cleanOther/Unknown/Declined -3.8793 14.4064 -0.269
## Pr(>|t|)
## (Intercept) 2.36e-11 ***
## ms_dx_condensedProgressive MS 1.95e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.17496
## demoEHR_Age 0.67869
## demoEHR_DiseaseDuration 0.50978
## race_ethnicity_cleanBlack Or African American 0.00104 **
## race_ethnicity_cleanHispanic or Latino 0.15875
## race_ethnicity_cleanWhite Not Hispanic 0.25791
## race_ethnicity_cleanOther/Unknown/Declined 0.78874
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21.85 on 54 degrees of freedom
## Multiple R-squared: 0.4843, Adjusted R-squared: 0.4079
## F-statistic: 6.338 on 8 and 54 DF, p-value: 8.902e-06hist(resid(home_matvel_dem_model))
home_matvel_dem_model_2 <- lm(PWS_velocitycmsecmean ~
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = home_uniqueid_df)
summary(home_matvel_dem_model_2)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = home_uniqueid_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -44.58 -15.01 -0.46 13.61 55.54
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 129.87748 26.40669 4.918
## ms_dx_condensedProgressive MS -34.14761 14.52045 -2.352
## ms_dx_condensedMS, Subtype Not Specified 22.10834 20.87100 1.059
## demoEHR_Age -0.03409 0.46211 -0.074
## demoEHR_DiseaseDuration -0.73408 0.80187 -0.915
## race_ethnicity_cleanBlack Or African American -64.58440 34.54653 -1.869
## race_ethnicity_cleanHispanic or Latino -23.21341 36.39556 -0.638
## race_ethnicity_cleanWhite Not Hispanic -13.90244 19.81309 -0.702
## race_ethnicity_cleanOther/Unknown/Declined -11.20032 24.26058 -0.462
## Pr(>|t|)
## (Intercept) 7.27e-05 ***
## ms_dx_condensedProgressive MS 0.0285 *
## ms_dx_condensedMS, Subtype Not Specified 0.3015
## demoEHR_Age 0.9419
## demoEHR_DiseaseDuration 0.3703
## race_ethnicity_cleanBlack Or African American 0.0756 .
## race_ethnicity_cleanHispanic or Latino 0.5305
## race_ethnicity_cleanWhite Not Hispanic 0.4906
## race_ethnicity_cleanOther/Unknown/Declined 0.6491
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 25.4 on 21 degrees of freedom
## Multiple R-squared: 0.4179, Adjusted R-squared: 0.1962
## F-statistic: 1.885 on 8 and 21 DF, p-value: 0.1168hist(resid(home_matvel_dem_model_2))
# Home videos - pws
hist(home_df$PWS_velocitycmsecmean)
ggplot(data = home_df, aes(x = log_delta_pix_h_rel_median_pose_hv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 4 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = stride_time_median_sec_pose_hv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 13 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = mean_cadence_step_per_min_pose_hv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 12 rows containing missing values (`geom_point()`).
ggplot(data = home_df, aes(x = stride_width_median_cm_pose_hv, PWS_velocitycmsecmean)) +
geom_point()## Warning: Removed 12 rows containing missing values (`geom_point()`).
## Univariate regressions - each video metric
all videos
metric_regression(home_df, PWS_velocitycmsecmean, log_delta_pix_h_rel_median_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 4 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -68.024 -16.891 5.468 16.520 45.090
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 139.620 8.162 17.106 < 2e-16 ***
## log_delta_pix_h_rel_median_pose_hv 25.141 5.844 4.302 6.72e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 24.29 on 57 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.2451, Adjusted R-squared: 0.2318
## F-statistic: 18.51 on 1 and 57 DF, p-value: 6.716e-05
## [1] "PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -39.622 -7.640 0.074 13.011 31.617
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 142.3431 16.4143 8.672
## log_delta_pix_h_rel_median_pose_hv 11.3031 5.0772 2.226
## demoEHR_Age 0.1397 0.2360 0.592
## demoEHR_DiseaseDuration 0.2060 0.4105 0.502
## ms_dx_condensedProgressive MS -39.3591 7.9626 -4.943
## ms_dx_condensedMS, Subtype Not Specified -5.5771 10.8498 -0.514
## race_ethnicity_cleanBlack Or African American -81.2997 21.6685 -3.752
## race_ethnicity_cleanHispanic or Latino -60.3534 18.9177 -3.190
## race_ethnicity_cleanWhite Not Hispanic -31.3722 11.2719 -2.783
## race_ethnicity_cleanOther/Unknown/Declined -20.3815 13.0623 -1.560
## clean_sexMale 29.2087 7.5998 3.843
## clean_sexNon-Binary 19.1975 13.4606 1.426
## Pr(>|t|)
## (Intercept) 2.58e-11 ***
## log_delta_pix_h_rel_median_pose_hv 0.030828 *
## demoEHR_Age 0.556642
## demoEHR_DiseaseDuration 0.618169
## ms_dx_condensedProgressive MS 1.02e-05 ***
## ms_dx_condensedMS, Subtype Not Specified 0.609640
## race_ethnicity_cleanBlack Or African American 0.000481 ***
## race_ethnicity_cleanHispanic or Latino 0.002532 **
## race_ethnicity_cleanWhite Not Hispanic 0.007727 **
## race_ethnicity_cleanOther/Unknown/Declined 0.125389
## clean_sexMale 0.000363 ***
## clean_sexNon-Binary 0.160421
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.56 on 47 degrees of freedom
## (4 observations deleted due to missingness)
## Multiple R-squared: 0.6747, Adjusted R-squared: 0.5985
## F-statistic: 8.861 on 11 and 47 DF, p-value: 2.983e-08
metric_regression(home_df, PWS_velocitycmsecmean, stride_time_median_sec_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 13 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ stride_time_median_sec_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -46.309 -20.619 2.342 19.588 38.570
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 186.63 16.10 11.592 1.62e-15 ***
## stride_time_median_sec_pose_hv -67.65 13.30 -5.088 5.97e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 23.31 on 48 degrees of freedom
## (13 observations deleted due to missingness)
## Multiple R-squared: 0.3504, Adjusted R-squared: 0.3369
## F-statistic: 25.89 on 1 and 48 DF, p-value: 5.965e-06
## [1] "PWS_velocitycmsecmean ~ stride_time_median_sec_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -34.464 -13.405 0.722 15.687 29.795
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 160.1095 31.2348 5.126 7.92e-06
## stride_time_median_sec_pose_hv -23.3736 22.9108 -1.020 0.31377
## demoEHR_Age 0.1376 0.3046 0.452 0.65402
## demoEHR_DiseaseDuration -0.1370 0.4739 -0.289 0.77403
## ms_dx_condensedProgressive MS -34.8772 17.9607 -1.942 0.05922
## ms_dx_condensedMS, Subtype Not Specified 3.6553 12.5589 0.291 0.77251
## race_ethnicity_cleanHispanic or Latino -55.2931 22.2496 -2.485 0.01723
## race_ethnicity_cleanWhite Not Hispanic -35.1292 15.2223 -2.308 0.02627
## race_ethnicity_cleanOther/Unknown/Declined -18.7907 15.7768 -1.191 0.24066
## clean_sexMale 28.9573 9.7313 2.976 0.00494
##
## (Intercept) ***
## stride_time_median_sec_pose_hv
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS .
## ms_dx_condensedMS, Subtype Not Specified
## race_ethnicity_cleanHispanic or Latino *
## race_ethnicity_cleanWhite Not Hispanic *
## race_ethnicity_cleanOther/Unknown/Declined
## clean_sexMale **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.27 on 40 degrees of freedom
## (13 observations deleted due to missingness)
## Multiple R-squared: 0.6298, Adjusted R-squared: 0.5465
## F-statistic: 7.562 on 9 and 40 DF, p-value: 2.276e-06
metric_regression(home_df, PWS_velocitycmsecmean, mean_cadence_step_per_min_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 12 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ mean_cadence_step_per_min_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -63.449 -18.164 7.176 19.931 41.250
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 63.7198 22.1577 2.876 0.00595 **
## mean_cadence_step_per_min_pose_hv 0.4228 0.2129 1.986 0.05264 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 27.84 on 49 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.0745, Adjusted R-squared: 0.05561
## F-statistic: 3.944 on 1 and 49 DF, p-value: 0.05264
## [1] "PWS_velocitycmsecmean ~ mean_cadence_step_per_min_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -34.860 -13.445 -0.673 14.720 29.959
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 151.1000 25.8511 5.845 7.19e-07
## mean_cadence_step_per_min_pose_hv -0.1668 0.1845 -0.904 0.37105
## demoEHR_Age 0.1525 0.2872 0.531 0.59831
## demoEHR_DiseaseDuration -0.0473 0.4602 -0.103 0.91863
## ms_dx_condensedProgressive MS -56.6810 11.0208 -5.143 7.07e-06
## ms_dx_condensedMS, Subtype Not Specified 0.1384 11.9538 0.012 0.99082
## race_ethnicity_cleanHispanic or Latino -57.7120 21.9668 -2.627 0.01205
## race_ethnicity_cleanWhite Not Hispanic -34.9418 15.0535 -2.321 0.02533
## race_ethnicity_cleanOther/Unknown/Declined -20.6330 15.6255 -1.320 0.19400
## clean_sexMale 27.6721 8.8391 3.131 0.00321
##
## (Intercept) ***
## mean_cadence_step_per_min_pose_hv
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS ***
## ms_dx_condensedMS, Subtype Not Specified
## race_ethnicity_cleanHispanic or Latino *
## race_ethnicity_cleanWhite Not Hispanic *
## race_ethnicity_cleanOther/Unknown/Declined
## clean_sexMale **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.1 on 41 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.6357, Adjusted R-squared: 0.5557
## F-statistic: 7.949 on 9 and 41 DF, p-value: 1.109e-06
metric_regression(home_df, PWS_velocitycmsecmean, stride_width_median_cm_pose_hv)## [1] "Data Frame: home_df"## Warning: Removed 12 rows containing missing values (`geom_point()`).
## [1] "PWS_velocitycmsecmean ~ stride_width_median_cm_pose_hv"
##
## Call:
## lm(formula = as.formula(outcome_predictor_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -67.022 -17.917 4.952 20.778 36.575
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 165.805 19.453 8.523 3.06e-11 ***
## stride_width_median_cm_pose_hv -4.627 1.503 -3.077 0.00341 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.49 on 49 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.162, Adjusted R-squared: 0.1449
## F-statistic: 9.47 on 1 and 49 DF, p-value: 0.003414
## [1] "PWS_velocitycmsecmean ~ stride_width_median_cm_pose_hv + demoEHR_Age + demoEHR_DiseaseDuration + \n ms_dx_condensed + \n race_ethnicity_clean + \n clean_sex"
##
## Call:
## lm(formula = as.formula(full_formula_string), data = data_frame)
##
## Residuals:
## Min 1Q Median 3Q Max
## -42.207 -11.526 1.933 14.318 28.690
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 157.1320 20.8368 7.541 2.87e-09
## stride_width_median_cm_pose_hv -2.2176 1.2391 -1.790 0.08090
## demoEHR_Age 0.2344 0.2808 0.835 0.40881
## demoEHR_DiseaseDuration -0.1056 0.4454 -0.237 0.81370
## ms_dx_condensedProgressive MS -44.4757 9.3356 -4.764 2.39e-05
## ms_dx_condensedMS, Subtype Not Specified 6.1075 11.9402 0.512 0.61174
## race_ethnicity_cleanHispanic or Latino -52.9084 21.4785 -2.463 0.01805
## race_ethnicity_cleanWhite Not Hispanic -36.0332 14.6224 -2.464 0.01801
## race_ethnicity_cleanOther/Unknown/Declined -20.5413 15.1766 -1.353 0.18332
## clean_sexMale 29.0007 8.5349 3.398 0.00152
##
## (Intercept) ***
## stride_width_median_cm_pose_hv .
## demoEHR_Age
## demoEHR_DiseaseDuration
## ms_dx_condensedProgressive MS ***
## ms_dx_condensedMS, Subtype Not Specified
## race_ethnicity_cleanHispanic or Latino *
## race_ethnicity_cleanWhite Not Hispanic *
## race_ethnicity_cleanOther/Unknown/Declined
## clean_sexMale **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.57 on 41 degrees of freedom
## (12 observations deleted due to missingness)
## Multiple R-squared: 0.6554, Adjusted R-squared: 0.5797
## F-statistic: 8.663 on 9 and 41 DF, p-value: 3.925e-07
### Multivariate regression Metrics only
# Metrics only - not including double support/stance measures, too many missing
# confounding +
home_matvel_multivar_model <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv,
data = home_df)
summary(home_matvel_multivar_model)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -44.785 -15.884 0.673 14.749 38.539
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 217.21836 41.01864 5.296 3.61e-06 ***
## log_delta_pix_h_rel_median_pose_hv 30.36748 10.52959 2.884 0.00606 **
## stride_time_median_sec_pose_hv -36.34131 19.91545 -1.825 0.07483 .
## mean_cadence_step_per_min_pose_hv -0.07951 0.22508 -0.353 0.72560
## stride_width_median_cm_pose_hv -1.82893 1.40931 -1.298 0.20114
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.84 on 44 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.5073, Adjusted R-squared: 0.4625
## F-statistic: 11.32 on 4 and 44 DF, p-value: 2.1e-06hist(resid(home_matvel_multivar_model))
# PWS interaction *
home_matvel_multivar_model_2 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv *
stride_time_median_sec_pose_hv *
mean_cadence_step_per_min_pose_hv *
stride_width_median_cm_pose_hv,
data = home_df)
summary(home_matvel_multivar_model_2)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv *
## stride_time_median_sec_pose_hv * mean_cadence_step_per_min_pose_hv *
## stride_width_median_cm_pose_hv, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -31.174 -9.053 0.587 11.557 32.855
##
## Coefficients:
## Estimate
## (Intercept) 1739.6382
## log_delta_pix_h_rel_median_pose_hv 3006.9649
## stride_time_median_sec_pose_hv 239.6412
## mean_cadence_step_per_min_pose_hv -8.1459
## stride_width_median_cm_pose_hv 7.9417
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv -1521.6538
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv -27.9368
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv -8.5989
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv -147.9248
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv -126.6266
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.6585
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 14.0448
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 54.0092
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 1.3684
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 1.7265
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.4769
## Std. Error
## (Intercept) 2953.2033
## log_delta_pix_h_rel_median_pose_hv 2554.3622
## stride_time_median_sec_pose_hv 2637.5459
## mean_cadence_step_per_min_pose_hv 34.7329
## stride_width_median_cm_pose_hv 191.2697
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv 2112.0019
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv 30.2273
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 32.4043
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv 164.3328
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 170.4884
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 2.2957
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 26.4519
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 132.4251
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 2.0017
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 2.1575
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 1.7381
## t value
## (Intercept) 0.589
## log_delta_pix_h_rel_median_pose_hv 1.177
## stride_time_median_sec_pose_hv 0.091
## mean_cadence_step_per_min_pose_hv -0.235
## stride_width_median_cm_pose_hv 0.042
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv -0.720
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv -0.924
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv -0.265
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv -0.900
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv -0.743
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.287
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.531
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 0.408
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.684
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.800
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv -0.274
## Pr(>|t|)
## (Intercept) 0.560
## log_delta_pix_h_rel_median_pose_hv 0.248
## stride_time_median_sec_pose_hv 0.928
## mean_cadence_step_per_min_pose_hv 0.816
## stride_width_median_cm_pose_hv 0.967
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv 0.476
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv 0.362
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.792
## log_delta_pix_h_rel_median_pose_hv:stride_width_median_cm_pose_hv 0.375
## stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 0.463
## mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.776
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv 0.599
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:stride_width_median_cm_pose_hv 0.686
## log_delta_pix_h_rel_median_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.499
## stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.429
## log_delta_pix_h_rel_median_pose_hv:stride_time_median_sec_pose_hv:mean_cadence_step_per_min_pose_hv:stride_width_median_cm_pose_hv 0.785
##
## Residual standard error: 18.31 on 33 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.7148, Adjusted R-squared: 0.5851
## F-statistic: 5.513 on 15 and 33 DF, p-value: 2.178e-05hist(resid(home_matvel_multivar_model_2))
# Metrics + disease and demographic info
# add MS subtype
home_matvel_multivar_model_3 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed,
data = home_df)
summary(home_matvel_multivar_model_3)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -43.66 -13.11 -0.56 13.79 38.14
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 178.5746 42.5981 4.192 0.000139
## log_delta_pix_h_rel_median_pose_hv 19.5787 10.7021 1.829 0.074441
## stride_time_median_sec_pose_hv 3.9832 25.2979 0.157 0.875642
## mean_cadence_step_per_min_pose_hv -0.1467 0.2131 -0.689 0.494917
## stride_width_median_cm_pose_hv -2.6326 1.4142 -1.862 0.069666
## ms_dx_condensedProgressive MS -44.1841 18.1056 -2.440 0.018972
## ms_dx_condensedMS, Subtype Not Specified 11.8323 11.2474 1.052 0.298814
##
## (Intercept) ***
## log_delta_pix_h_rel_median_pose_hv .
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv .
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.59 on 42 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.5844, Adjusted R-squared: 0.5251
## F-statistic: 9.844 on 6 and 42 DF, p-value: 9.041e-07hist(resid(home_matvel_multivar_model_3))
# add age
home_matvel_multivar_model_4 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age,
data = home_df)
summary(home_matvel_multivar_model_4)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed + demoEHR_Age,
## data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -40.391 -13.190 1.107 15.944 38.803
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 193.3250 44.7742 4.318 9.74e-05
## log_delta_pix_h_rel_median_pose_hv 17.5002 10.8671 1.610 0.1150
## stride_time_median_sec_pose_hv -4.0523 26.3843 -0.154 0.8787
## mean_cadence_step_per_min_pose_hv -0.1557 0.2129 -0.731 0.4687
## stride_width_median_cm_pose_hv -2.2229 1.4645 -1.518 0.1367
## ms_dx_condensedProgressive MS -39.7174 18.5690 -2.139 0.0384
## ms_dx_condensedMS, Subtype Not Specified 16.4033 12.0371 1.363 0.1804
## demoEHR_Age -0.2525 0.2391 -1.056 0.2971
##
## (Intercept) ***
## log_delta_pix_h_rel_median_pose_hv
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.56 on 41 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.5954, Adjusted R-squared: 0.5264
## F-statistic: 8.62 on 7 and 41 DF, p-value: 1.825e-06hist(resid(home_matvel_multivar_model_4))
# add disease duration
home_matvel_multivar_model_5 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration,
data = home_df)
summary(home_matvel_multivar_model_5)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -39.656 -13.530 0.656 15.458 37.868
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 194.6680 45.5440 4.274 0.000115
## log_delta_pix_h_rel_median_pose_hv 17.6598 11.0065 1.604 0.116475
## stride_time_median_sec_pose_hv -4.7910 26.8185 -0.179 0.859117
## mean_cadence_step_per_min_pose_hv -0.1608 0.2161 -0.744 0.461329
## stride_width_median_cm_pose_hv -2.2169 1.4815 -1.496 0.142390
## ms_dx_condensedProgressive MS -39.9088 18.7942 -2.123 0.039951
## ms_dx_condensedMS, Subtype Not Specified 16.6309 12.2025 1.363 0.180534
## demoEHR_Age -0.2233 0.2637 -0.847 0.402120
## demoEHR_DiseaseDuration -0.1169 0.4207 -0.278 0.782602
##
## (Intercept) ***
## log_delta_pix_h_rel_median_pose_hv
## stride_time_median_sec_pose_hv
## mean_cadence_step_per_min_pose_hv
## stride_width_median_cm_pose_hv
## ms_dx_condensedProgressive MS *
## ms_dx_condensedMS, Subtype Not Specified
## demoEHR_Age
## demoEHR_DiseaseDuration
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.78 on 40 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.5962, Adjusted R-squared: 0.5155
## F-statistic: 7.383 on 8 and 40 DF, p-value: 5.496e-06hist(resid(home_matvel_multivar_model_5))
# add race and ethnicity
home_matvel_multivar_model_6 <- lm(PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
stride_time_median_sec_pose_hv +
mean_cadence_step_per_min_pose_hv +
stride_width_median_cm_pose_hv +
ms_dx_condensed +
demoEHR_Age +
demoEHR_DiseaseDuration +
race_ethnicity_clean,
data = home_df)
summary(home_matvel_multivar_model_6)##
## Call:
## lm(formula = PWS_velocitycmsecmean ~ log_delta_pix_h_rel_median_pose_hv +
## stride_time_median_sec_pose_hv + mean_cadence_step_per_min_pose_hv +
## stride_width_median_cm_pose_hv + ms_dx_condensed + demoEHR_Age +
## demoEHR_DiseaseDuration + race_ethnicity_clean, data = home_df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -41.954 -11.525 1.007 13.271 40.271
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 198.11475 45.35758 4.368
## log_delta_pix_h_rel_median_pose_hv 23.50946 10.84331 2.168
## stride_time_median_sec_pose_hv 2.23808 25.94949 0.086
## mean_cadence_step_per_min_pose_hv -0.13102 0.21082 -0.621
## stride_width_median_cm_pose_hv -1.70982 1.45321 -1.177
## ms_dx_condensedProgressive MS -40.98916 18.23476 -2.248
## ms_dx_condensedMS, Subtype Not Specified 9.05339 12.36775 0.732
## demoEHR_Age -0.02147 0.28319 -0.076
## demoEHR_DiseaseDuration 0.15047 0.47339 0.318
## race_ethnicity_cleanHispanic or Latino -55.28232 22.30640 -2.478
## race_ethnicity_cleanWhite Not Hispanic -27.65538 14.74643 -1.875
## race_ethnicity_cleanOther/Unknown/Declined -25.00977 15.72431 -1.591
## Pr(>|t|)
## (Intercept) 9.74e-05 ***
## log_delta_pix_h_rel_median_pose_hv 0.0367 *
## stride_time_median_sec_pose_hv 0.9317
## mean_cadence_step_per_min_pose_hv 0.5381
## stride_width_median_cm_pose_hv 0.2469
## ms_dx_condensedProgressive MS 0.0306 *
## ms_dx_condensedMS, Subtype Not Specified 0.4688
## demoEHR_Age 0.9400
## demoEHR_DiseaseDuration 0.7524
## race_ethnicity_cleanHispanic or Latino 0.0179 *
## race_ethnicity_cleanWhite Not Hispanic 0.0686 .
## race_ethnicity_cleanOther/Unknown/Declined 0.1202
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.02 on 37 degrees of freedom
## (14 observations deleted due to missingness)
## Multiple R-squared: 0.6548, Adjusted R-squared: 0.5521
## F-statistic: 6.38 on 11 and 37 DF, p-value: 8.558e-06hist(resid(home_matvel_multivar_model_6))
Home Video significant association of PWS velocity with
log delta pix h rel median after adjusting for dem and disease info
Linear Regression –> FW velocity (check distribution)
FW Zeno videos